Full text of "The New Cambridge Medieval History"

See other formats

The period of the tenth and early eleventh centuries was crucial in the
formation of Europe, much of whose political geography and larger-
scale divisions began to take shape at this time. It was also an era of great
fragmentation, and hence of differences which have been magnified by
modern national historiographical traditions. The international team of
authors in this volume of The New Cambridge Medieval History reflects these
varying traditions, and provides an authoritative survey of the period in
its own terms.

The volume is divided into three sections. The first covers common
themes and topics such as the economy, government, and religious cul-
tural and intellectual life. The second is devoted to the kingdoms and
principalities which had emerged within the area of the former
Carolingian empire, as well as the ‘honorary Carolingian’ region of
England. The final section deals with the emergent principalities of
eastern Europe and the new and established empires and statelets of the
Mediterranean world.

Cambridge Histories Online © Cambridge University Press, 2008

The New Cambridge Medieval History

EDITORIAL BOARD

David Abulafi a Rosamond McKitterick
Martin Brett Edward Powell
Simon Keynes Jonathan Shepard
Peter Linehan Peter Spuff ord

Volume hi c. 900— c. 1024

Cambridge Histories Online © Cambridge University Press, 2008

THE NEW
CAMBRIDGE
MEDIEVAL HISTORY

l olume in c. 900— c. 1024

EDITED BY

TIMOTHY REUTER

Professor of Medieval History
University of Southampton

gjjj Cambridge

UNIVERSITY PRESS

Cambridge Histories Online © Cambridge University Press, 2008

CAMBRIDGE UNIVERSITY PRESS

Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, Sao Paulo

Cambridge University Press
The Edinburgh Building, Cambridge cb 2 2RU, UK

Published in the United States of America by Cambridge University Press, New York

www.cambridge.org

Information on this title: www.cambridge.org/97805 21 364478

This book is in copyright. Subject to statutory exception
and to the provisionsof relevant collective licensing agreements,
no reproduction of any part may take place without
the written permission of Cambridge University Press.

First published 1999
Reprinted 2006

Printed in the United Kingdom at the University Press, Cambridge

A. catalogue record for this book is available from the British Library

Library of Congress Cataloguing in Publication data

ISBN- 1 3 978-0-521-36447-8 paperback
ISBN- 10 0-521-46447-7 hardback

Cambridge University Press has no responsibility for the persistence or accuracy of URLs
for external or third-party internet websites referred to in this book, and does not guarantee
that any content on such websites is, or will remain, accurate or appropriate.

Cambridge Histories Online © Cambridge University Press, 2008

CONTENTS

List of maps

pagey

List of plates

List of contributors

xiii

Preface

List of abbreviations

xviii

Introduction: reading the tenth century

TIMOTHY REUTER

PART i: GENERAL THEMES

Rural economy and country life

ROBERT FOSSIER

2 7

Merchants, markets and towns

PETER JOHANEK

Rulers and government

JANET L. NELSON

The Church

ROSAMOND MCKITTERICK

I ?o

Monasticism: the first wave of reform

JOACHIM WOLLASCH

!6 3

Intellectual life
CLAUDIO LEONARDI

Artists and patrons

HENRY MAYR-HARTING

212

vii

Cambridge Histories Online © Cambridge University Press, 2008

Vlll

CONTENTS

PART Ii: POST-CAROLINGIAN EUROPE

9 The Ottomans as kings and emperors 233

ECKHARD MULLER-MERTENS

i o Saxony and the Elbe Slavs in the tenth century 267

GERD ALTHOFF

1 1 Bavaria in the tenth and early eleventh centuries 29 3

HERWIG WOLFRAM

1 2 Lotharingia 310

MICHEL PARI S SE

13 Burgundy and Provence, 879-1032 328

CONSTANCE BRITTAIN BOUCHARD

14 The kingdom of Italy 346

GIUSEPPE SERGI

15 West Francia: the kingdom 372

JEAN DUNBABIN

1 6 West Francia: the northern principalities 398

DAVID BATES

17 Western Francia: the southern principalities 420

MICHEL ZIMMERMANN

1 8 England, c. 900- 1016 456

SIMON KEYNES

PART hi: NON-CAROLINGIAN EUROPE

19 European Russia, c. 500— c. 1050 487

THOMAS S. NOONAN

20 Bohemia and Poland: two examples of successful Slavonic

state-formation 514

JERZY STRZELCZYK

21 Hungary 536

KORNEL BAKAY

22 Byzantium in equilibrium, 886—944 5 5 3

JONATHAN SHEPARD

Cambridge Histories Online © Cambridge University Press, 2008

Contents

2 3 Bulgaria: the other Balkan ‘empire’ 5 67

JONATHAN SHEPARD

24 Byzantium expanding, 944-1025 586

JONATHAN SHEPARD

2 5 Byzantium and the West 60 5

JONATHAN SHEPARD

26 Southern Italy in die tenth century 624

G. A. LOUD

27 Sicily and al-Andalus under Muslim rule 646

HUGH KENNEDY

28 The Spanish kingdoms 670

ROGER COLLINS

Appendix: ruler and genealogical tables 69 3

hist of primary sources 718

Bibliography of secondary works arranged by chapter 737

Index 812

Cambridge Histories Online © Cambridge University Press, 2008

MAPS

1 Urban settlements and emporia in the Scandinavian and Baltic

regions 65

2 Archbishoprics and bishoprics in the early eleventh century 1 3 2—3

3 Monastic centres in tire tenth and early eleventh centuries 1 64

4 Germany 234— 5

5 Lotharingia 3 1 1

6 Burgundy and Provence 329

7 The kingdom of Italy, c. 1000 347

8 The kingdom of France, c. 1000 373

9 England in the tenth century 457

I o European Russia in tire ninth and tenth centuries 489

I I Poland, Bohemia and Hungary 5 1 5

1 2 Bulgaria in the tenth century 5 69

13 Byzantium in 1025 588—9

14 Southern Italy 625

15 The Spanish peninsula, c. 1 000 671

Cambridge Histories Online © Cambridge University Press, 2008

PLATES

frontispiece

London, Victoria and Albert Museum, the Basilewski ivory situla , c. 980
(photo © Victoria and Albert Museum)

between pages 228 and 229

1 Canigou, church of St Martin du Canigou, early eleventh century

2 Gernrode, church of St Cyriacus, late tenth century (photo: Bildarchiv
Foto Marburg)

3 Cologne, church of St Pantaleon, interior of Westwork, late tenth
century

4 Oxford, Bodleian Library, page of the Winchester Troper, with quern
queritis, late tenth century (photo: © Bodleian Library, Oxford)

5 Essen Minster, sheath of a ceremonial sword, late tenth century

6 Cologne cathedral, Gero crucifix, c. 970

7 Essen minster, processional cross of the Abbess Matilda of Essen,
971—82 (photo: Bildarchiv Foto Marburg)

8 Essen minster, interior of Westwork, late tenth century (photo:

Bildarchiv Foto Marburg)

9 Essen minster, crown of Pthe boy Otto III, late tenth century (photo:
Bildarchiv Foto Marburg)

10 Trier cathedral, reliquary of St Andrew’s foot, 977— 93 (photo: Bildarchiv
Foto Marburg)

1 1 Conques, statue of St Faith, tenth century (photo: Bildarchiv Foto
Marburg)

1 2 Essen minster, statue of the ‘golden Madonna’, tenth century (photo:
Bildarchiv Foto Marburg)

13 Paris, Bibliotheque Nationale, Ms. lat 5(2), fo. 173, first Bible of St
Martial, Limoges, showing initial letter to the Acts of the Apostles, c. 1000
(photo: Bibliotheque Nationale, Paris)

Cambridge Histories Online © Cambridge University Press, 2008

xii LIST OF PLATES

14 Ivea, Biblioteca, Capitolare, Ms. 86, psalter of Bishop Warmund of Ivrea,
showing the standing figure of Habakkuk, c. 1000

1 5 Ivrea, Biblioteca Capitolare, Ms. 85 , sacramentary of bishop Warmund of
Ivrea, showing a grieving woman at the graveside, c. 1000

1 6 Wolfenbiittel, Stadtarchiv, marriage roll of the empress Theophanu, 972
(photo: Bildarchiv Foto Marburg

1 7 Aachen minster, Aachen Gospels, showing the emperor Otto III seated
in majesty, c. 996 (photo © Ann Munchow, Aachen)

1 8 Florence, Museo Bargello, Byzantine ivory of seated Christ ascending to
heaven, tenth century

19 Cividale del Friuli, Museo Nazionale Ms. 136,60. 17, psalter of
Archbishop Egbert of Trier, showing Ruodprecht presenting the book to
the archbishop, 977—93

20 Bamberg, Staatsbibliothek, Ms. Lit. 5 3, fo. zv, pontifical made at Seeon,
showing the solemn entry into church of the emperor Flenry II flanked
by two bishops, 1014—24 (photo: Staatsbibliothek, Bamberg)

Cambridge Histories Online © Cambridge University Press, 2008

CONTRIBUTORS

gerd althoff, Professor of Medieval History, University of Munster
david bates , Professor of Medieval History, University of Glasgow
Constance bouchard, Professor of Medieval History, Kenyon College, Ohio
roger collins, Edinburgh

jean dunbabin ,Fellow of St. Anne’s College, University of Oxford
robert fossier, Professor of History, University of Paris I
peter johanek, Professor of Regional History, University of Munster
hugh Kennedy, Reader in History, University of St Andrews
simon keynes, Felloiv of Trinity College, University of Cambridge
claudio leonardi, Professor of Latin Philology, University of Florence
g. a. loud, Reader in 'History, University of Leeds

henry mayr-harting, Regius Professor of Ecclesiastical History, University of
Oxford

rosamond mckitterick, Professor of Medieval History, University of
Cambridge

eckhard muller-mertens, Professor of History Emeritus, Humboldt
University of Berlin

janet l. wesson. Professor of History, King’s College, London
t. s. noonan, Professor of History, University of Minnesota
michel parisse, Professor of Medieval History, University of Paris 1 7 II
timotliy reuter, Professor of Medieval History, University of Southampton
giuseppe sergi, Professor of Medieval History, University of Turin
Jonathan shep ard, Fellow of Peterhouse, University of Cambridge
jerzy strzelczyk, Professor of History, University of Poynari
herwig wolfram, Director, Institut fiir osterreichische Geschichtsforschung,
University of Vienna

joachim wollasch, Professor of Medieval History Emeritus, University of
Mimster

michael zimmermann, Professor of Medieval History, University of Paris

xiii

Cambridge Histories Online © Cambridge University Press, 2008

PREFACE

Timothy Reuter

No one can be more aware than a volume editor of the difficulties inherent in
the project of a New Cambridge Medieval History, not least the argument that all
such projects belong to a positivistic attitude to knowledge which has now
rightly passed from the stage. Had the intention simply been to make a better
job of providing a ‘definitive account’ of this particular section of the past
than was done under the editorship of J. P. Whitney when volume 1 1 1 of the
old Cambridge Medieval History, subtitled ‘Germany and the Western Empire’,
was published in 1922, die project would indeed seem problematic. But peri-
odic stock-takings are both important and necessary, especially given that
approaches to the early medieval past have changed so fundamentally in the
last seventy years. They allow a group of scholars to set out for a wider audi-
ence the current state of play in their own areas of specialisation, and so to
provide students, teachers and the general public with a set of accounts of the
subject which have all been produced at much the same time and to much the
same set of instructions. The result may no doubt date, though slowly, but it is
in any case no longer expected to do anything else. If the framework is still, as it
was in the early years of this century, drat of political history, it is a political
history conceived more broadly, and, it is to be hoped, more readably, than was
current in die 1 920s. My introductory chapter and those of the other contribu-
tors to the opening thematic section set out some of the links between political
history and other ways of practising the discipline.

The division around 1024 between this volume and its twin successors,
inherited from the earlier Cambridge Medieval History, obviously has no immedi-
ate significance except for German, Italian and (more or less) Byzantine
history, and it has been appropriately modified for the chapters on other topics.
Both it and its substitutes here are divisions conceived essentially in terms of
political history, but diis has die positive advantage of not having to plump for
either of the current rival datings on offer for the Great Medieval Shift: that
from the ancient world to the medieval world (or from slavery to feudalism)

Cambridge Histories Online © Cambridge University Press, 2008

XVI

PREFACE

around 1000; or drat from ‘archaic society’ to the ‘Old European Order’
around 1050. More is said of these and other interpretative schemata in the
introductory chapter.

An intellectual climate more relativistic than that which prevailed in die time
of Acton, Whitney and Tanner has had die advantage for the editor that he has
felt little pressure to harmonise interpretations and interpretative styles
between contributions, though he hopes that there are few if any remaining
discrepancies in respect of ‘facts’. Indeed, it is a positive advantage that the
reader should become more aware of die great range of approaches to early
medieval history currently being practised in this country, on die continent and
in North America. It is for this reason that the team of contributors is a fairly
international one rather than being restricted to Anglophone historians. To
have followed the latter course would have had many advantages, but would
have risked presenting die reader with a greater appearance of homogeneity in
current approaches to the subject than really exists. Intellectual stock-taking
should take account not only of what is currently thought but of how and why
it has come to be so thought, and in particular should emphasise rather than
conceal die differences between national historiographical traditions. In the
introduction I have attempted to set out some of die implications of these
traditions and explore their strengths and weaknesses.

The volume is arranged in three parts. The chapters in the opening section
cover themes not easily or sensibly divided up geographically. The following
section has nine chapters on the polities which emerged after the break-up of
die Carolingian empire, and also includes the chapter on England, which was
institutionally, culturally and politically an important part of the post-
Carolingian order. The final section covers non-Carolingian Europe (including
Byzantium and die Islamic polities within Europe), with the chapters arranged
from north-east to south-west. In order to avoid too many mini-chapters,
some responsibilities have been divided between this volume and its prede-
cessor. Volume II contains accounts of the histories of the Scandinavian
peninsula and of the Celtic regions which extend into the tenth and early
eleventh centuries. The present volume has a full account of Russian history
from its earliest stages to 1054; die chapter planned on Jews and Jewish life in
western Europe from 700 to 1050 fell victim to the death of a contributor and
die impossibility of finding a replacement who could undertake to deliver
within a reasonable space of time. Originally planned chapters on lordship and
on warfare suffered similar fates; a little of the ground which would have been
covered in these chapters is touched on in my introductory chapter, which is
for that reason longer than it otherwise might have been.

Each chapter has its own bibliography of secondary sources (including
works not referred to in the footnotes), but references to primary sources are

Cambridge Histories Online © Cambridge University Press, 2008

Preface xvii

made by short tide to the consolidated bibliography of primary sources. The
spelling of place-names follows the conventions in use by Cambridge
University Press. The spelling of medieval personal names is inevitably in part
a matter of prejudice and habit. The editor has on the whole preferred an
Anglophone, more international and less anachronistic mode of spelling:
Radulf/Rudolf rather than Ralph or Raoul, Odo rather than Eudes, Henry
rather than Heinrich, Enrico or Henri. The results may on occasion be unfa-
miliar, but do at least have die advantage drat diey do not give to tenth-century
people who in fact bore die same name spellings of that name which vary arbi-
trarily according to whereabouts in twentieth-century Europe they happen to
have been studied. Traditional forms like Raoul and Eudes are cross-refer-
enced in the index. Technical terms have largely been left in their Latin (or ver-
nacular) forms, and they are explained on their first occurrence.

In the course of an enterprise of diis kind one incurs many debts. I owe
thanks to all my contributors, especially to diose who responded to what were
often very belated proposals for changes and cuts widi consideration and cour-
tesy, and also to those contributors who did meet the original deadline for
delivery punctually and then found themselves waiting in limbo. Most, though
certainly not all, of the materials for the volume were ready at the time of my
move to Southampton in 1994, and although the contributors have kept their
bibliographies up to date they have made only minor changes to their texts.
The delays since 1994 have had a number of causes: illness; pressure of other
university duties; and not least the publication of other volumes in die series,
which have set precedents and so forced me to redo some editorial work I had
thought finished and to undertake other work I had not anticipated having to
do. The readers of this volume will not suffer as a result of the delays, but some
of die contributors have, and I am grateful to them for their forbearance.

I am very grateful to Dr Sarah Hamilton (Southampton) and Dr Eleanor
Screen (Peterhouse, Cambridge) for their assistance in checking references and
bibliographies in the final stages of preparation. My special thanks go to Jinty
Nelson, Jonathan Shepard and Chris Wickham for their friendship and for their
freely granted advice and support on both the intellectual and the psychological
problems involved in planning the volume and in dealing with contributors.
During the whole period of preparation Rosamond McKitterick and I have
exchanged much advice and information on our respective volumes, and I
should like to thank her here for this and for much-needed support at various
difficult points in the gestation of the volume. Last but not least I must thank
William Davies and the staff at Cambridge University Press most warmly for the
help they have given at all stages, and for their patience in awaiting delivery.

Timothy Reuter

Cambridge Histories Online © Cambridge University Press, 2008

ABBREVIATIONS

AASS Acta Sanctorum quotquot toto orbe coluntur, ed .

J. Bollandus etal., Antwerp and Brussels
(1634-)

Adalbert, Reginonis Continuatio Adalbert of St Maximin, Reginonis Continuation

ed. F. Kurze, Regino of Priim, Chronicon,

Adam of Bremen, Gesta

Adhemar, Chronicon

AfD
AHP
AHR
AKG
An. Boll.
Annales ESC

A Q

ASC

ASE

BAR

BEC

BHL

pp. 154-79

Adam of Bremen, Gesta Hammaburgensis
ecclesiae pontificum , ed. B. Schmeidler, MGH
SRG11, Hanover (1917)

Adhemar of Chabannes, Chronicon , ed.

J. Chavanon, Ademarde Chabannes, Chronique
publiee d’apres les manuscrits, Paris (1897)

Archiv fur Diplomatik

Archivum historiae pontificae

American Historical Review

Archiv furKulturgeschichte

Analecta Bollandiana

Annales: Economies, Societes, Civilisations

Ausgewahlte Quellen %urdeutschen Geschichte des

Mittelalters (Freiherr-von-Stein- Geddchtnis-Ausgabe)

Anglo-Saxon Chronicle, trans. Whitelock, EHD,

pp. 145-245

Anglo-Saxon England

British Archaeological Reports

Bibliotheque de IE cole des Chartes

Bibliotheca hagiographica latina, subsidia

hagiographicaw, Brussels (1 898-1901),

Supplementum, subsidia hagiographica x 1 1 ,

Brussels (191 1); Novum supplementum, subsidia

hagiographica Brussels (1986)

xviii

Cambridge Histories Online © Cambridge University Press, 2008

List of abbreviations

xix

Bib. Mun.

BLMS

BUGS

BN lat., BN n.a. lat.

BSl

B JZ

By^bulg

CBA

CCCM

CCM

CCSL

CFHB

Clm

D(D)

DBI

DCI

DC II

DChS

DHI

D H II

Bibliotheque Municipale

London, British Library manuscript

Byzantine and Modern Greek Studies

Paris, Bibliotheque Nationale, manuscrit latin;

nouvelles acquisitions latines

Bygantinoslavica

Bygantion

Bytymtinobulgarica

Bygantinische Zeitschrift

Council for British Archaeology

Corpus Christianorum, Continuatio

mediavalis, Turnhout (i 966-)

Corpus consuetudinem monasticarum,
ed. K. Hallinger, Siegburg (1963-)

Corpus Christianorum, series latina,

Turnhout, (19 5 2-)

Corpus fontium historiae Byzantinae
Munich, Bayerische Staatsbibliothek, Codex
Latinus Monacensis

Diploma(ta), cited by number in the following
editions:

Berengar I, king of Italy, Diplomata,
ed. L. Schiaparelli, I diplomi di Berengario I
(sec. IX— X) (Fonti per la storia d’ltalia 35),
Rome (1903)

Conrad I, king of east Francia, Diplomata,
ed. T. Sickel, Die Urkunden Konrad I., Heinrich I.
und Otto I. (MGH Dip. regum 1), Hanover
(1879-84)

Conrad II, emperor, Diplomata, ed. H.

Bresslau, Die Urkunden Konrads II. (MGH
Dip. regum iv) Berlin (1909)

Charles the Simple, king of west Francia, Acta,
ed. P. Lauer, Recueil des actes de Charles III le
Sirrrple, roi de France, 893—923, Paris (1949)

Henry I, king of east Francia, Diplomata, ed.

T. Sickel, Die Urkunden Konrad I, Heinrich I. und
Otto I. (MGH Dip. regum 1), Hanover
(1879-84)

Henry II, king of east Francia and emperor,
Diplomata, ed. H. Bresslau, H. Bloch,

Cambridge Histories Online © Cambridge University Press, 2008

LIST OF ABBREVIATIONS

D Hugh

DLIV

DLC

D LG

DLo

D Lothar

DO I

DO II
D O III
DRa

DRoI

R. Holtzmann, M. Meyer and H. Wibel,

(MGH Dip. regum III) , Hanover
(1900-3)

Hugh, king of Italy, Diplomata, ed. L.
Schiaparelli, I diplomi di Ugo e diLotario, di
Berengario II e diAdalberto (secolo X) (Fond per la
storia d’ltalia 38), Rome (1924)

Louis IV, king of west Francia ,Acta, ed. P.
Lauer, Recueil des actes de Louis IX ] roi de France

(936-9H), Paris (1914)

Louis the Child, king of east Francia,
Diplomata, ed. T. Schieffer, Die Urkunden
Zmntibolds und Ludwigs des Kindes (MGH Dip.
Germ, iv), Berlin (i960)

Louis (the German), king of east Francia,
Diplomata, ed. P. Kehr, Ludwig des Deutschen,
Karlmanns und Ludwigs des Jungeren Die Urkunden
(MGH Dip. Germ. 1), Berlin (1932—4)

Lodiar, king of west Francia, Acta, ed. L.
Halphen and F. Lot, Recueil des actes de Lothaire
et Louis X( rois de France (944—98-/), Paris
(1908)

Lodiar, king of Italy, Diplomata, ed. L.
Schiaparelli, I diplomi di Ugo e diLotario, di
Berengario II e diAdalberto (secolo X) (Fond per la
storia d’ltalia 38), Rome (1924)

Otto I, king of east Francia, Diplomata, ed.

T. Sickel, Die Urkunden Konrad I, Heinrich I,
und Otto I. (MGH Dip. regum 1), 2 vols.,
Hanover (1 879- 84)

Otto II, Diplomata, ed. T. Sickel, Die Urkunden
Otto des II. (MGH Dip. regum 1 1 . 1), Hanover
(1888)

Otto III, Diplomata, ed. T. Sickel, Die Urkunden
Otto des III. (MGH Dip. regum 1 1 .2), Hanover

(r 893)

Radulf (Raoul), king of west Francia, Acta, ed.
R.-H. Bautier and J. Dufour, Recueil des actes de
Robert Ier et de Raoul, rois de France, 922—446,

Paris (1978)

Robert I, king of west Francia, Acta, ed.

Cambridge Histories Online © Cambridge University Press, 2008

List of abbreviations

xxi

D Ro II

DAI

DOP

DOT

EHD

EHR

EME

ep(pf

Eparch

Flodoard, Annates

Flodoard, HRE

FmaSt

fol.

FSI

Fulbert, Ep(p).
Gerbert, Ep(p).

R.-H. Bautier and J. Dufour, Recueil des actes de
Robert Ieret de Raoul, rois de France, 922—936,

Paris (1978)

Robert II, king of west Francia,Hrfe, ed.

W. M. N ewman, Catalogue des actes de Robert II,
roi de France, Paris (1937)

Deutsches Archiv fi-ir Erforschung des Mittelalters
Constantine VII Porphyrogenitus, De
administrando imperio, ed. and trans.

G. Moravcsik and R. J. H. Jenkins (CFHB

1 = Dumbarton Oaks Texts 1), Washington,
DC (1967)

Constantine VII Porphyrogenitus, De
cerimoniis aulae byxantinae, ed. 1 . 1 . Reiske,

2 vols., Bonn (1 829).

Dumbarton Oaks Papers
Dumbarton Oaks Texts

Dorothy Whitelock (ed.), English Historical
Documents 1, c. 900—1042, 2nd edn (London,

079)

English Historical Review
Early Medieval Europe
epistola(e)

The Book of The Eparch, ed. and trans.

J. Koder, Das Eparchenbuch Leons des Weisen
(CFHB 33, Series Vindobonensis), Vienna
099 1 )

Flodoard, Annales, ed. P. Lauer, Les annales de
Flodoard publiees d’apres les manuscrits, Paris
( I 9°5)

Flodoard, Historia Remensis ecclesiae, ed.

M. Stratmann, MGH SS xxxvi, Hanover
099 8 )>

Friihmittelalterliche Studien
folio

Fond per la storia d’ltalia (Instituto storico per
il medio evo) (1887-)

The Letters and Poems of Fulbert of Chartres, ed.
and trans. F. Behrends, Oxford (1 976)

Gerbert of Aurillac, Epistolae, ed. F. Weigle,
Die Briefsammlung Gerberts von Reims ( MGH Die

Cambridge Histories Online © Cambridge University Press, 2008

XXII

LIST OF ABBREVIATIONS

HJb

JEccH

JMH

ldm<Ppnnd,Antapodosis
Liudprand, Historic!

Liudprand, Relatio

Mansi

MGH

Cap.

Cap. episc.
Cone.

Const.

Dip. Germ.

Brief e der deutschen Kaiseryeit 1 1) , Weimar
(1966)

Historisches Jahrbuch

Historische Zeitschrift

Journal of Ecclesiastical History

P. Jaffe, Regesta pontificum romanorum, 2nd edn,

ed. S. Loewenfeld, with F. Kaltenbrunner and

P. Ewald, Leipzig (1885-8)

Journal of Medieval History
Liudprand of Cremona, Antapodosis, ed.

J. Becker, Lindprandi opera (MGH SRGxia),
Hanover (19 1 5), pp. 1-158
Liudprand of Cremona, Liber de rebus gestis
Ottonis Magni imperatoris, ed.J. Becker,
Lindprandi opera (MGHSRG xli), Hanover
( I 9 1 5 ), PP- 1 59 — 75

Liudprand of Cremona, Relatio de kgatione
Constantinopolitana , ed.J. Becker, Lindprandi
opera (MGH SRGxli), Hanover (191 5),
pp. 175-212
LeMoyenAge

J. D. Mansi, Sacrorum Conciliorum nova et
amplissima eolketio, Florence and Venice
( I 757 ~ 98)

Monumenta Germaniae Historica, with subseries:
Auctores antiquissimi, 1 5 vols., Berlin
(1877-1919)

Capita laria. Legum sectio 11, Capitularia regum
Francorum, ed. A. Boretius and V. Krause,

2 vols., Hanover (1883-97)

Capita la episcoporum, ed. P. Brommer, Hanover
(1984)

Concilia. Legum sectio in, Concilia , n, ed.

A. Werminghoff, Hanover (1 906-8); in, ed.

W Hartmann, Hanover (1984); iv, ed.

W Hartmann, Hanover (1998)

Constitutiones et acta publica imperatorum et regum
inde ab a. dccccxi usque ad a. mcxccvii
(gn—iipy), ed. L. Weiland, Hanover (1893)

Dip lorn ata regum Germaniae ex stirpe Karolinorum:
Die Urkunden der deutschen Karolinger 1 , ed.

Cambridge Histories Online © Cambridge University Press, 2008

MGH (cont.)
Dip. Kar.
Dip. regum

Epp.

Epp. sel.

Fontes

Form.

Leges nat. Germ.

Lib. mem.
Nee. Germ.
Poet.

SRG

SRL

SRM

List of abbreviations xxiii

P. Kehr, Berlin (1932—4); 11, ed. P. Kehr, Berlin
(1936-7); in, ed. P. Kehr, Berlin (1956);
iv, ed. T. Schieffer, Berlin i960)

Diplomata Karolinorum: Die Urkunden der
Karo linger 1 and hi, ed. E. Miihlbacher and
T. Schieffer, Hanover (1893-1908)

Diplomata regum et imperatorum Germaniae: Die
Urkunden der deutschen Konige und Kaiser 1 , ed.

T. Sickel, Hanover (1879-84); 11.1, ed.

T. Sickel, Hanover ( 1 8 8 8) ; 1 1 . 2 , ed. T. Sickel,
Hanover (1 893); in, ed. H. Bresslau, H. Bloch
and R. Holtzmann, Hanover (1900-3); iv, ed.
H. Bresslau, Berlin (1 909)

Epistolae iii-vin (= Epistolae Merovingici et
Karolini aevi , Hanover (1892-1939)

Epistolae selectae in usum scbolarum, 5 vols.,
Hanover (1887- 91)

Fontes iuris Germanici antiqui in usum scbolarum ex
Monumentis Germaniae Historicis separatim editi,

1 3 vols., Hanover (1 909-86)

Formulae Memvingici et Karolini aevi , ed.

K. Zeumer, Legum section , Hanover (1886)
Leges nationum Germanicarum , ed. K. Zeumer
( Lex Visigoth oruni)-, L. R. de Salis ( Leges
Burgundionuni)-, F. Beyerle and R. Buchner ( Lex
Ribuaria)-, K. A. Eckhardt {Pactus legis Salicae
and Lex Salicd)-, E. von Schwind (, Lex
Baiivariorum), 6 vols. in 1 1 parts, Hanover
(1892—1969)

Libri memoriaks, and Libri memoriales et
Necrologia nova series, Hanover (1 979- )
Necrologia Germaniae , 5 vols. and Suppl.
Hanover (1 886-1 920)

Poetae Latini aevi Carolini, ed. E. Diimmler,

L. Traube, P. von Winterfeld and K. Strecker,

4 vols., Hanover (1881-99)

Scrip tores rerum Germanicarum in usum scbolarum
separatim editi, 6 3 vols. , Hanover (1871-1987)
Scrip tores rerum Langobardicarum et Italicarum saec.
1 1 —IX, ed. G. Waitz, Hanover (1 878)

Scriptores rerum Merovingicarum, ed. B. Krusch

Cambridge Histories Online © Cambridge University Press, 2008

xxiv

LIST OF ABBREVIATIONS

MGH (cont.)
MIOG

MMS

N.F.

n.s.

QFIAB

Radulf Glaber, Historiae
RB

Regino, Chronicon

RHEF

RHF

RhVjb

Richer, Historiae

RISS

and W Levison, 7 vols., Hanover (1 885-1920)
Scriptores (in Folio), 30 vols., Hanover
(1824-1924)

Mitteilungen des Instituts fur Osterreichische
Geschichtsforschung (1922-1944, Mitteilungen des
Osterreichischen Instituts fur Geschichtsforschung)
MIOG, Ergan^ungsband
Munstersche Mittelalterschriften
Manuscript

NeuesArchiv der Geselleschaftfurdltere deutsche
Geschichtskunde, continued as Deutsches Archiv
fiir Erforschung des Mittelalters
Neue Folge
nova series, new series
Pa tro logia e cursus completus, series graeca, ed.

J.-P. Migne, 161 vols. (Paris, 1857-66)
Patrologiae cursus completus, series latina , ed.

J.-P. Migne, 221 vols., Paris (1841-64)

Quellen und Forschungen aus italienischen Archiven
und Bibliothehen

Radulf Glaber, Historiarum libri quinque , ed.
with English trans. J. France, Oxford (1989)
Revue Benedictine

Regino of Priim, Chronicon, ed. F. Kurze,
Reginonis abbatis Prumiensis Chronicon cum
continuatione Treverensi, MGHSRG l, Hanover
( i 8 9 °)

Revue d’Histoire de lEglise de France
Academie des Inscriptions et Belles-Lettres, Recueil
des historiens des Gaules et de la France, series in
folio, eds. M. Bouquet and M.-J.-J. Brial,
revised by L. Delisle, 19 vols., Paris
(1869—80)

Rheinische l lerteljahrsbldtter
Richer, Historiae, ed. and trans. R. Latouche,
Richer, Histoire de France (888—99;) (Classiques
de l’histoire de France au moyen age), 2 vols.,
Paris (1930, 1937; repr. i960, 1964)

Rerum italicarum scriptores, ed. L. A. Muratori,

25 vols., Milan (1723-51); new edn,

G. Carducci and V. Fiorini, Citta di Castello
and Bologna (1900-)

Cambridge Histories Online © Cambridge University Press, 2008

XXV

List of abbreviations

s.a.

Sawyer

Settimane

Skylitzes, Synopsis

StMGBO

Thietmar, Chronicon

TRHS

Vat. (lat.; pal. lat.; reg. lat.)

VSWG

VuF

Widukind, Res gestae Saxonicae

Wipo, Gesta

ZRG

sub anno

P. H. Sawyer, Anglo-Saxon Charters:

A Handlist, London (1968)

Settimane di Studio del Centro Italiano di Studi
sull’alto medioevo (Spoleto 1955—)

John Skylitzes, Synopsis historiarum, ed.

I. Thurn (CFHB 5, Series Berolinensis), Berlin
and New York (1973)

Studi Mediaevali

Studien und Mitteilungen %ur Geschichte des
Benediktiner- Ordens und seiner Zweige
Thietmar of Merseburg, Chronicon, ed.

R. Holtzmann {MGH SRGN.S. ix, Berlin
(i935)

Transactions of the Royal Historical Society
Bibliotheca Apostolica Vaticana, MS (latinus;
palatinus latinus; reginensis latinus)
l lerteljahresschrft fur Wirtschaftsgeschichte
Vortrage und Forschungen, herausgegeben
vom Konstanzer Arbeitskreis fur
mittelalterliche Geschichte
Widukindi monachi Corbeiensis rerum gestarum
Saxonicarum libri III, ed. P. Hirsch and H.-E.
Lohmann (MGH SRG lx , Hanover (1935)
Wipo, Gesta Chuonradi, ed. H. Bresslau, Wiponis
opera {MGH SRG lxi ), Hanover (1915),
pp. 3-62

Zeitschrift der Savigny-Stiftungfur Rechtsgeschichte

GermanistischeAbteilung

Kanonistische Abteilung

Cambridge Histories Online © Cambridge University Press, 2008

CHAPTER I

INTRODUCTION: READING
THE TENTH CENTURY

Timothy Reuter

the present volume covers a period in European history best described as
the ‘long tenth century’, stretching from die 890s through to around 1020/ 30.
Though diis volume covers Byzantine history of die period and also Islamic
history so far as it impinges on European territory, the emphasis in this intro-
duction will be largely on what was or would become the Latin west. I shall try
to sketch what currently seem the main concerns of historians working on the
period and what are generally seen as its salient features, though any such
attempt will probably date far faster than the substantive chapters which
follow The ways in which historians make and have made sense of the period
as a whole have been determined by a range of inputs. Before we can look at
the general trends which are currently held to characterise the period (and the
extent to which they actually do) we need to examine these inputs. The most
important of them is the nature, real and perceived, of the available source-
materials. But two others are almost as important. The first comes from the tra-
ditional and non-traditional interpretative schemata and periodisations which
the community of professional scholars has brought to bear. The second,
perhaps even more important, is tire fact that the members of this community
for the most part work and have worked within specific historiographical tradi-
tions.

It is widely held that the long tenth century is a period more lacking in
sources and reliable and precise information on ‘what actually happened’ than
any other period of post- Roman European history, with the exception perhaps
of the seventh century. It is not just the very evident brutality of much of the
period that has caused it to be termed a ‘dark century’ ( dunkles Jahrhundert) or an
‘obscure age’ ( secolo oscuro), or an ‘iron age’ (with die overtone, so chilling for
modern professional scholars, that words and thoughts are silenced in the face
of armed force). 1 It is also the difficulty historians often encounter, for

1 See Zimmermann (1971), pp. 1 5—21, on the history of these terms; Lestocquoy (1947), White (1955)

and Lopez (1962) are early attempts at re-evaluating the period as a conscious reaction against them.

Cambridge Histories Online © Cambridge University Press, 2008

TIMOTHY REUTER

example, when trying to establish precise sequences of events or office-
holders. At least in parts of the post-Carolingian core of Europe there seems
to have been a decline in pragmatic literacy and a reversion to oral and symbolic
means of communication. As we shall see, this was by no means a universal
feature of the long tenth century; but to the extent that it did really exist it
meant that human interaction often took forms which have inevitably left rela-
tively fewer traces in the written record, and those often indirect and difficult to
interpret.

Nevertheless, notions of a dark or obscure or ‘iron’ age are problematic.
Though they go back a long way, they exercised their most formative influence
during the period when a Rankean primacy of political history still dominated
medievalists’ consciousnesses. When there is at most one substantial narrative
dealing with the high politics of a region, writing about ‘what actually hap-
pened’ seems even more difficult and uncertain than it is in any case, and the
results thus dark or obscure. Many regions of Europe are in this position for
most of the long tenth century: east Frankish/German history is unusual in
having the accounts of Widukind of Corvey, Liudprand of Cremona and
Adalbert of St Maximin running in parallel for much of die middle third of the
tenth century.

Even this dearth of narratives is a difficulty found mainly in the west, Latin
and Islamic, rather than the east, where the tenth century is no more obscure
than any other period of Byzantine history and rather less than some. Outside
the Mediterranean world there are indeed regions for which we have virtually
no contemporary narratives at all. The emergent realms of Rus', Hungary,
Bohemia and Poland, naturally, as well as the Scandinavian kingdoms, have no
contemporary indigenous accounts, only later, mythologising origin histories:
the Tak of Bygone Years or Russian Primary Chronicle for Rus'; the late twelfth-
century Anonymus and later derivatives like Simon de Keza and the Chronicon
pictum for Hungarian history; the early twelfth-century court writers, Cosmas
of Prague and Gallus Anonymus, for Bohemian and Polish history; Saxo
Grammaticus, Heimskringla and its precursors for Scandinavian history. The
savage positivist source-criticism of the late nineteenth and early twentieth
centuries has left few historians willing to use such works as ‘primary sources’
except in a state of cautious desperation or for the citation of an occasional
phrase to add rhetorical colour. Even when it is evident that their authors must
have drawn on earlier works now lost to us, it is normally impossible to tell pre-
cisely where they are doing this, while the analysis of these works as later repre-
sentations of an earlier past has in many cases barely begun. Once the
information offered by these high-medieval versions of earlier pasts is seen as
the product of later construction rather than the echo of past reality, the polit-
ical history of these regions has to be written in a much more tentative and

Cambridge Histories Online © Cambridge University Press, 2008

Introduction: reading the tenth century

uncertain fashion, drawing mainly on casual and largely decontextualised frag-
ments of information found in narratives from the Frankish, Anglo-Saxon and
Byzantine world and in Arabic and Jewish travellers’ tales. Some parts of
western Europe are almost as badly placed, most notably the kingdom of
Burgundy and the principalities of Catalonia and Toulouse, at least as far as any
reconstruction of histoire evenementielle is concerned: few European rulers of
any period can have left as little trace in the record after reigning for nearly sixty
years as has Conrad the Pacific of Burgundy.

Yet tire long tenth century is also an age of great historians, writers who offer
rich and juicy texts with a wide narrative sweep and much significant detail:
Widukind of Corvey, Adalbert of Magdeburg and Thietmar of Merseburg
working in Saxony; Flodoard and Richer in Rheims; Dudo of Saint-Quentin in
Normandy; Adhemar of Chabannes and Radulf Glaber in central France;
Liudprand of Cremona in Italy (and north of the Alps); Benedict of Soracte in
Rome; Sampiro in Leon. Some sections and some versions of the enigmatic
complex known collectively as the Anglo-Saxon Chronicle , notably the strange
compilation by the ealdorman dEthelwold written around 980, would also
qualify. There are also impressive works of more local compass, such as the
Lotharingian episcopal gesta, or Flodoard’s lengthy and archivally based history
of the church of Rheims. Most important of all, and not only for the sheer
bulk of what survives, is the large corpus of saints’ lives and miracle-
collections from this period: it was a golden age of hagiographic production.

Traditional attitudes, however, are slow to change. Modern medievalists’
relationship with ‘hagiography’ is revealed by the fact that whereas almost all
the major ‘historiographical’ works of the period are available in good modern
editions, most ‘hagiography’ still has to be consulted in old and often very inad-
equate editions. A nineteenth-century distinction between historians, who deal
in facts, and hagiographers, who deal in fictions, was perhaps appropriate to an
era of scholarship in which it was important to begin by establishing the who,
the what, the where and the when, all matters on which ‘hagiographic’ texts are
often imprecise or inaccurate. But it now needs to be transcended: it is by no
means clear that the distinction reflects anything significant about the inten-
tions and practices of tenth-century authors: many ‘historians’ also wrote
‘hagiography’. 2

Yet few even of those conventionally thought of as historians rather than
hagiographers have left us straightforward and unproblematic texts. The acid-
bath of positivist source-criticism may have dissolved die later mydiologising
histories of the European periphery almost completely, but it has also left the
smoodi surfaces of writers like Widukind, Richer and Dudo deeply pitted, so

2 Lifshit2 (1994).

Cambridge Histories Online © Cambridge University Press, 2008

TIMOTHY REUTER

much so that Martin Lintzel could write about die ‘problem of trudi in the
tenth century’ (meaning the problem of having any confidence at all in the rela-
tion between our surviving accounts and the course of an increasingly inac-
cessible past reality ‘out there’), and more recendy Carlrichard Briihl has felt
able to dismiss both Widukind and Richer as romanciers? Few historians at the
end of the twentiedi century are still willing to offer this kind of robust empiri-
cism without qualms; but though the aspects of these sources problematised
by Lintzel and Briihl are not the only ones, they are real enough, for elements
of saga, of epic, of the preacher’s exemplum, of folk-tale, seem to greet us on
many pages of these works, and they will rarely submit to a straightforward
positivist unpacking of their meaning . 4

Historians of a positivist frame of mind have traditionally contrasted the
uncertain and subjective information derived from narratives with the firmer
data to be won from record evidence, which in this period means from charters.
Many series of royal diplomata from diis period now exist in complete and
satisfactory modern editions: diose issued by or in the name of the rulers of
east Francia/ Germany, of Burgundy, of Hungary and of Italy are available
complete, and those of the west Frankish rulers almost so, while as far as sur-
viving papal letters and privileges are concerned it is for this period alone that
we possess a comprehensive edition of everything surviving . 5 Even for those
regions where the picture is still incomplete — Anglo-Saxon England, the
Spanish peninsula, Byzantium - the gaps are being filled. Below that level the
picture is less favourable. Although die period is characterised by the exercise of
‘quasi-regal’ power by figures with less dian royal status - archbishops, bishops,
dukes, margraves - the charters they issued were not numerous, and in most
regions have hardly begun to be collected in modern editions ; 6 an exception is
the collection of the placita of die kingdom of Italy, accounts of judicial deci-
sions given by a court president acting (or ostensibly acting) in the ruler’s name . 7

The bulk of non-royal charter material surviving from this period consists
of what we would nowadays think of as either conveyancing records or
accounts of dispute setdement. Normally such documents offer a miniature
narrative of a conveyance or settlement with a list of diose present at die trans-
action; in many areas of northern Europe they were treated, so far as we can
tell, as a mere record of the transaction widi no inherent legal force, though
both England and Italy show diat this did not have to be the case. It is precisely
during the period covered by this volume that the narratives in many parts of

3 Lintzel (1956); Briihl (1990), pp. 465— 7, 589— 93. 4 Reuter (1994).

5 Zimmermann, H. (ed.), Papsturkunden 896—1046.

6 Kienast (1968) provides a convenient guide to the charters produced for secular princes; there is a

complete edition for Normandy in Recueil des actes des dues de Normandie.

7 Manaresi, C. (ed.), Iplaciti del ‘Regnum Italiae’.

Cambridge Histories Online © Cambridge University Press, 2008

Introduction: reading the tenth century

Europe, especially in France, become less miniature and more detailed, and it
has indeed been argued that such loquacity has misled historians into drinking
that the tilings they describe in such detail were really new around the millen-
nium rather than simply coming to be recorded for the first time . 8 Both their
geographical distribution and the quality of the editions they have received are
very uneven. The archives of the Mediterranean regions - Italy, both nordr and
south, and parts of Spain (especially Catalonia and Castile) - are very full, if
not always very fully known or exploited. In nordrern Europe such collections
of material as have survived have normally done so in the form of cartularies
put together by religious institutions, often in die century and a half after die
period covered by this volume, when such institutions were taking steps to put
their property ownership and administration on a more ordered and rational
basis, and so to arrange selected and edited versions of their archives in book
form. Large and unmediated archival deposits are rare, the large tenth- and
eleventh-century archives of Cluny being an unusual exception . 9 In particular,
many of the north European centres active in producing archival material in
the eighth and ninth centuries, from Redon to St Gallen, either ceased to do so
altogedier in the tenth century or else did so at a greatly reduced rate.

Little of this material has been edited both comprehensively and recently.
Nor has its nature always been properly appreciated by historians. The history
of diplomatic has been one of a preoccupation with distinguishing the genuine
from the false. The question of authenticity is an appropriate and important
point from which to start when dealing with royal and papal charters, because
such documents, at least in theory, were in themselves adequate to guarantee
the claims contained in them, and this made them worth forging, both at the
time and later. But it does not go far enough, even for them. Every charter tells
a story, and even if we can establish that tire charter is indeed what it purports
to be, the authenticity of the charter in a formal legal sense is in itself no guar-
antee of the authenticity or completeness or meaningfulness in a historical
sense of the story which it tells. Most such stories are indeed manifestly incom-
plete, and historians have barely begun to study the narrative strategies of
charter-writers and of those who commissioned their activities. This is all the
more significant with the advent, already noted, of a much more garrulous
style of charter-writing, including plaints ( querimoniae ) and concords ( convenien -
tiae) which set out the whole history of a dispute. The fact that these miniature
histories are found embedded in what look like legal documents does not make
them any less subjective or their interpretation any less problematic.

In some, though not all parts of Latin Europe there was a temporary down-
turn in charter production in the early part of this period, though the view of the

Barthelemy (1992a). 9 Recueil des chartes de I’abbaye de Cluny.

Cambridge Histories Online © Cambridge University Press, 2008

TIMOTHY REUTER

period as an ‘obscure age’ has itself obscured the fact that this downturn was
reversed almost everywhere by the later tenth century, to be followed by steady
growth. But there was a quite genuine and long-lasting downturn in legislative
activity almost everywhere in Latin Europe; it was one of the most evident con-
trasts between the Latin west on the one hand and Byzantine or Islamic political
culture on the other, for those few contemporaries who were familiar with
both. 10 For most of the west during this period little or no legislation survives,
even in those regions where rulers appear to have been powerful and impressive
figures, and this is not to be attributed to large-scale losses of what once existed.
The Carolingian capitulary tradition had virtually died out by the end of the ninth
century (after 884 m west Francia, after 898 in Italy, after 852 m east Francia). The
Ottomans and their entourages knew what capitularies were, but confined them-
selves to very occasional ad hoc edicts. 11 Collections of Carolingian capitularies,
notably that of Ansegis, continued to be copied in the tenth and early eleventh
centuries, both in west and in east Francia in particular, but it is far from clear
what use might have been made of such manuscripts in practical life. 12 Anglo-
Saxon England is the great western European exception to the tenth-century leg-
islative drought; here, collections of Carolingian capitularies transmitted from
the continent provided some of the inspiration which enabled the kingdom to
catch up with, absorb and develop the lessons of Carolingian government in a
long series of law-codes, notably those of fiEthelstan, iEthelred and Cnut. 13 Paler
forms of imitation of the Carolingians can be seen in the laws of Stephen of
Hungary from the early eleventh century. 14 The Byzantine development was, as
one might expect, smoother and more continuous: the tenth-century rulers con-
tinued to legislate as a matter of course, without break or decline. 15

The church also legislated less: councils, where they did meet, were more
likely to leave only protocols of judicial decisions or charters solemnised by the
fortuitous presence of numerous imposing witnesses than they were to
produce legislation in the form of canons. 16 Equally, tire great Carolingian tra-
dition of episcopal capitularies had comparatively weak echoes in the practice
of tenth-century bishops. 17 This picture of inactivity is particularly true of the

10 See Nelson’s analysis of John of Gorze’s account of his visit to the Cordovan court, below, pp. 1 26—8.

11 MGH Const 1, no. 8, p. 17; D H II 370.

12 Mordek (1995); Ansegis, Collectio capitularium , ed. Schmitz, pp. 1 89—90.

13 Edited in Die Geset^e dersl ngelsachsen, ed. Liebermann; on the Carolingian sources for such legislation

see Wormald (1978), pp. 71—4. 14 Stephen, King of Hungary, Lam.

15 See Shepard, below, pp. 5 5 3—4; on the contrast with the west in this respect see Leyser (1994b), pp.
160— 1.

16 This is the conclusion of Schroder (1980) for west Francia; the situation elsewhere was similar if less
extreme.

17 Capitula episcoporum in contains a few tenth-century specimens; the overall distribution of texts and
manuscripts is to be surveyed in vol iv, which has not yet appeared.

Cambridge Histories Online © Cambridge University Press, 2008

Introduction: reading the tenth century

early tenth century; from around 950 onwards there was something of a recov-
ery. Although this recovery was hardly a rapid one anywhere, the great
sequence of reforming councils initiated by Leo IX’s councils at Rheims and
Mainz in 1 049 was not preceded by a long legislative drought in tire way that tire
otherwise comparable revival of conciliar activity in the early Carolingian
period had been. 18 Our picture is still an imperfect one, for though such secular
laws as have survived, in Byzantium and in the west, have generally been well
edited, conciliar legislation is only now receiving the attention it deserves. 19 In
particular, we lack a comprehensive edition of tire texts produced by those
councils at which the ‘legislation’ of the Peace and Truce of God movements
was promulgated. 20 But we also lack a modern edition of almost any of the col-
lections of canon law regularly used in the long tenth century, or of the great
collection produced at tire end of it by Burchard of Worms, which largely
superseded these earlier collections. 21

Almost all of the surviving letter-collections of the period (and not many
tenth-century letters have been preserved outside collections) can be seen in a
context of canon law. It is not an accident that the most important ones are
associated with important reforming clerics - Rather of Verona and Liege,
Gerbert of Rheims, Fulbert of Chartres, Dunstan of Canterbury - and that
they contain many letters dealing with practical matters of church law. 22 Letters
should not be seen in this context alone, however. The impulse to preserve
them in collections, which would become stronger and more widespread in the
course of the eleventh and twelfth centuries, was not simply a product of
the period’s concern with memoria and of a desire to preserve die memory of
the people with whom they were associated. It also stemmed from die need for
models to be used in the training of clerics: significandy, Dunstan, Gerbert and
Fulbert were teachers as well as lawyers. The Latin poetry of the period was
also located in this rhetorical-didactic tradition: an art of the schools rather
than of the court, which it had been at least to some extent in die preceding
period. 23 Here again we have a contrast between the Latin west and the court-
centred cultures of Byzantium and Islam.

As with the earlier medieval centuries, one feels diat the material remains of

18 Hartmann (1989) pp. 47— 50.

19 Concilia aevi Saxonici 916—1001, 1: 916—961', for commentary see Schroder (1980), Vollrath (1985), Wolter
(1988), and the chapters in the forthcoming History of Medieval Canon Law edited by Wilfried
Hartmann and Kenneth Pennington.

20 See Hoffmann (1 964) for details of the printed sources; much of the manuscript work remains to be
done.

21 Hoffmann and Pokorny (1 99 1) is now the starting point for any work on Burchard’s collection.

22 Rather of Verona, Epistolae ; Gerbert of Aurillac, Epistolae', Fulbert of Chartres, The Letters and Poems',

Memorials of Saint Dunstan, pp. 354—438. The connection is most evident in the case of Fulbert: see,
e.g., epp. 28, 36, 56, 71. 23 Godman (1987).

Cambridge Histories Online © Cambridge University Press, 2008

TIMOTHY REUTER

the long tenth century ought to have made more impact on historians’ con-
sciousnesses and interpretations than in practice they have done. Excavation
has played a major part in reshaping post-Carolingian urban history, not least
through die very detailed investigation of Viking York and Dublin; Peter
Johanek’s chapter shows how dais has affected our view of the period. Our
view of post-Carolingian settlement patterns owes in general much less to
archaeology: this is certainly true of villages, which, as Robert Fossier argues
below (in common with many other though by no means all scholars), first
start to take on definitive form and permanent location in this period. It is
perhaps less true of the dwellings of the dominant aristocratic strata of post-
Carolingian society, also seen as ‘settling down’ in the course of the long tenth
century, but although the development of dae aristocratic dwelling, often a for-
tified site, has been extensively studied and has been linked to shifts in family
structure in this period, we are still far from having a clear view of where and
how the non-urban aristocracies of northern Europe lived . 24 Historians of the
tenth century should undoubtedly pay more attention to archaeology than they
have, though the absence of substantial syntheses and the gaps in the publica-
tion of excavations as well as the divergences between national archaeological
traditions (even more marked than die historiographical divergences to be
examined shortly) will continue to make this difficult in the foreseeable future.

Some kinds of material remains have escaped historians’ general neglect of
non-written sources, most notably those traditionally studied by art historians:
painting, sculpture, goldsmithery and ivorywork, architecture. The study of
manuscripts, both as material objects and as repositories of images, has
received at least as much attention as the study of the written sources of the
period. So have die surviving remains of metalwork and wood- and ivory-
carvings, in the form of book-covers and other carved panels, of liturgical
combs, and above all of reliquaries and items of regalia. Much of this record is
lost, however, and some of its context is irrecoverable. Virtually no secular
buildings and very few ecclesiastical ones have survived unchanged and intact
from the tenth century. The wall-paintings and tapestries which once deco-
rated them, and which would probably have told us even more about the
culture and self-image of the period than do illuminated manuscripts, have
vanished almost without trace, except for an occasional survival like the church
of St George on the Reichenau with its almost intact cycle of wall-paintings.
Ecclesiastical vestments have survived in quite substantial numbers, but the
tapestries recording the deeds of kings and aristocrats are known only from a
handful of casual written references. Many of these kinds of material survival
have attracted the attention of cultural and political historians as well as of his-

24 See below, pp. 18 — 19 .

Cambridge Histories Online © Cambridge University Press, 2008

Introduction: reading the tenth century

torians of art, because they fall or can be seen as falling into the category of
‘signs of lordship and symbols of state’, to use a phrase invented by the
German medievalist Percy Ernst Schramm. Like their counterpart in written
sources, the (often anecdotalised) record of symbolic action, they have seemed
to offer a way in to the mindset of the period’s elites which might otherwise be
closed to us by the sheer inarticulacy of more direct evidence . 25

The source-materials available for the study of a period are far from defin-
ing the ways in which that period will be studied. Claudio Leonardi begins his
chapter on intellectual life by remarking that the era between the late
Carolingian scholars and litterati and die early scholastics of the later eleventh
century is often thought of either as post-Carolingian or as pre-Gregorian,
and is thus denied an identity of its own . 26 Analagous remarks could be made
about the prevailing interpretation of other aspects of the period. There is, of
course, some justification for such terminology and the interpretative sche-
mata which lie behind it. Much of tenth-century Europe — though hardly the
Byzantine and Islamic spheres - saw itself as in a sense post-Carolingian: it
simultaneously perpetuated and looked back nostalgically to an order once
glorious, now in decline. The heirs of die direct successor-states looked back
to a supposedly golden age of Frankish unity, which seemed all the more
golden for the absence of any clear and precise memories of it. Carolingian
nostalgia was at its strongest in regions where the Carolingians had been
largely absent, like the south of France, and it grew once real Carolingians
were no longer around: it was Otto III, not Otto I, who took the first steps
towards the canonisation of Charlemagne . 27 The post-Carolingian core of
Europe retained a residual sense of pan-Frankishness long after kingdoms
(not, as yet, nations), had started to develop their own sense of identity. In die
large arc to die north and east of the former Frankish empire, from England
through to Hungary, it was as much the written and unwritten myth of the
Carolingian polity as experience of the contemporary hegemonial power, the
Ottomans, that provided a model for development, whedier in the form of
imitation capitularies in the Wessex of Edgar and fEthelred or in the adapta-
tion of Lex Bahmariorum to serve as the basis for early Hungarian law. Equally,
although the ‘Gregorian’ and ‘pre-Gregorian’ terminology may have been sub-
jected to powerful attacks in recent years it can hardly be escaped altogether . 28
The apparent universality of the charges laid by the church reformers and his-
torians of the mid- and late elevendi century and echoed by historians of
the nineteenth and twentieth at least gives a degree of unification to our per-
ceptions of tenth- and early eleventh-century Europe, united by sin, by

25 For the work of Schramm see Bak (1973); for work on political ritual see Althoff (1990); Koziol

(1992); Althoff (1997). 26 Below, p.187.

27 Folz (1950), pp. 47—114; Remensnyder (1995). 28 Tellenbach (1985, 1993).

Cambridge Histories Online © Cambridge University Press, 2008

IO TIMOTHY REUTER

ecclesiastical abuse, and by attempts by a small radical minority to overcome
these failings.

Two odier models currendy offer broader versions of the divisions just
mentioned. Much German-language historiography — and formerly French
historiography as well, as witness Marc Bloch’s distinction between the first
and the second feudal age - sees the mid-eleventh century as having marked a
crucial change from an ‘archaic’ society to that ‘old European order’ which pre-
vailed from the late eleventh to the late eighteenth century. 29 This may be seen
as a more secular and sociological rewriting of die schema ‘pre-’ and ‘post-
Gregorian’: church reform was on this view merely symptomatic of more
general changes in the eleventh century towards greater rationality and greater
social differentiation. 30

An alternative view, which would stress political more than other kinds of
development, is to see the period as initiating, as far as Latin-speaking western
Europe is concerned, a very long era during which Europe would be shaped by
competing dynastically oriented territories, many of them the ancestors of the
modern nation-state, even though that term is hardly applicable to die tenth
century. Geoffrey Barraclough defined the long tenth century as the ‘crucible
of Europe’, the period in which large-scale supra-regional empires finally dis-
appeared, to be replaced by the smaller kingdoms familiar from later European
history. 31 Certainly much of Europe’s political geography can be seen to have
begun in this period, a fact which was taken as die basis of a large international
conference in 1968 on the ‘origins of nation-states’ in this period. 32 Yet even as
an interpretation of political history alone it fits some parts of Europe much
better than it does others. It clearly works well for the northern and eastern
parts of Europe, where present-day polities very evidently emerged from pre-
history in a recognisable form in die course of the tenth century. German
medieval historiography has also devoted much attention to the ‘beginnings of
German history’, which are now generally placed in the course of the long
tenth century rather than the ninth, even if they are no longer defined in terms
of a significant date like 91 1 or 919 or 93 b. 33

Yet it is German medievalists who have sought to establish the ‘beginnings
of French history’ and place them in die same period; 34 it is far less of a defin-
ing moment for French historians, for whom something recognisable as
France had already been around for some time by the tenth century. Indeed it is
in the French historiographical tradition that a quite opposite view has been
developed. Rather than the ‘birdi of Europe’ rhetoric, this offers the tenth

29 Brunner (1968); Gerhard (1981) For Bloch’s distinction see Bloch (1961), pp. 59—71 .

30 Murray (1978), esp pp. 25-137.

31 Barraclough (1976); cf. also the titles of Calmette (1941), Fossier (1982) and Fried (1991).

32 Manteuffel (1968). 33 Bruhl (1990); Ehlers (1994). 34 Ehlers (1985).

Cambridge Histories Online © Cambridge University Press, 2008

Introduction: reading the tenth century

1 1

century as the last century of an old order, one which was not merely post-
Carolingian, but post-Roman. The reasons which have been given for taking
such a view have varied. Some scholars have wanted to stress a continuity of
the late antique legal and political order through to the late tenth century . 35
Others, Marxisant or neo- Marxisant, have stressed an underlying shift in the
mode of production and hence the dominant social formation from slavery to
serfdom (and hence, in the Marxist sense of the terminology, from slavehold-
ing to feudalism ). 36 Others have seen die tenth century as ending in a new frag-
mentation ( encellulement ) of society, a world in which interaction at a distance
had almost ceased to exist, in which the horizon did not extend much beyond
the view from the castle wall . 37

With considerations like these we have already arrived at the third kind of
input mentioned at the outset, and it is not only for tire reasons just discussed
that the interpretative schemata on offer for tenth-century history depend on
the historiographical tradition in which a historian is working. There is a
common European tradition, but its regional variations are very marked. In
particular, the master narratives dominant in the various European countries
and regions mean that there is no comprehensive European consensus on
which aspects of tire period are to be seen as significant. To some extent there
is also a problem of language: both the technical terms and tire underlying con-
ceptual apparatus in use vary from national tradition to national tradition, and
there are as yet few guides to these which will allow the historian to carry out
reliable translation. It may well be that an increasing awareness of other tradi-
tions and of the work being done within them will create a more genuinely
European view of tenth-century history within the coming generation; some
of what we currently perceive as real differences in the past may turn out to be
mere differences of perception, the products of divergent terminology and
historiographical tradition.

It is noteworthy how many of tire periodisations and implicit or explicit
underlying models are drawn from French history, and in an English-lan-
guage history it is worth stressing the point. Not only have French medieval-
ists been given to offering such theories more than most; both the
Anglo-Norman and Anglo-Angevin connections of English medieval history
and tire foreign-language teaching traditions dominant in the Anglolexic
world have created a ‘Francocentric’ approach: French medieval history has

35 Durliat (i99o);Magnou-Nortier (1981, 1982, 1984); for a critique see Wickham (1993) The same peri-
odisation is found, more impressionistically justified, in Sullivan (1989).

36 Bois (1989); Bonnassie (1991).

37 Fossier (1982), pp. 288—601, esp. pp. 288—90; also below, pp. 45—53. For the relationship between
encellulement and incastellamento , its Italian relative, see the historiographical account in Wickham
(1986), pp. xxiii— xxvi; for critiques of the concept see Leyser (1994c) and Campbell (1990).

Cambridge Histories Online © Cambridge University Press, 2008

TIMOTHY REUTER

often been taken metonymicaUy in Britain and America for the whole of
tenth- and early eleventh-century Europe. More important still is the way in
which an impressive series of regional studies, beginning with and in many
cases inspired by Duby’s classic study of the Maconnais, have fleshed out in
often very substantial detail the transformation of various parts of France in
the post-Carolingian era . 38 We have a better picture of die tenth century on
the ground for west Francia than for any other part of Europe, not necessar-
ily because the supply of sources is inherendy superior, but because many of
its regions have been systematically studied in a way in which tenth-century
Bavaria or Umbria have not yet been (it would be possible to do so, and
indeed French historians have themselves exported the approach beyond the
boundaries of west Francia ). 39 This is, arguably, accident: the original Annales
idea of ‘total history’ has simply turned out to be more easy to realise by his-
torians of the high middle ages than by historians of later periods in the time
available for the production of theses. If this is so, it has been a very signifi-
cant accident.

The positions and traditions of Italian and Spanish medievalists show great
similarities. The tenth century is one of extreme localisation: meaningful gen-
eralisations about or general histories of the Italian or Spanish peninsulas are
difficult, if not impossible. Moreover, the master narratives of Italian and
Spanish historiography make the tenth century a period of marking time:
waiting for the communes, or for the reconquista , and so looking for the antece-
dents of these tilings. The tenth century hardly works for either Italy or Spain
as the end of an old or die beginning of a new era. Although it is possible to
talk about the first half of die tenth century as one in which Italy was ruled by
‘national’ kings, this is only acceptable nowadays when accompanied by a heavy
coating of inverted commas. Nor is the tenth century a significant one for
Spanish self-perception. On the one hand, the crucial period for the survival of
the kingdom of Leon-Asturias and its taking firm root was the ninth, not the
tenth century. On die other hand, Spanish political geography was not defini-
tively shaped until much later. Castile, which would ultimately play Wessex to
most of the rest of the peninsula, was still an insecure border region in this
period. There has also been much to do. Professional history-writing has not
been so long established or so well funded as in the lands north of Alps and
Pyrenees, and there is still an immense amount of positivist establish-the-facts
spadework to be done for this period. It is significant, therefore, that Italian
and Spanish historians have been heavily influenced in recent years by the con-
cerns of French medievalists. Two large and highly influential studies, those of

38 Duby (1952); most of the others are listed in Poly and Bournazel (1991), English translation, pp.

365—6.

39 E.g. Toubert (1973 a, 1973b); Bonnassie (1975, i976);Taviani-Carozzi (1991); Menand (1993).

Cambridge Histories Online © Cambridge University Press, 2008

Introduction: reading the tenth century 1 3

Pierre Toubert on Latium and of Pierre Bonnassie on Catalonia, have been
particularly important in setting agendas. 40

As is explained in the preface, die present volume is ordered by reference to
the tenth century’s Carolingian past: the chapters on the ‘post-Carolingian
core’ are grouped before those on what from this point of view was the periph-
ery, diough neither die Byzantines nor the Islamic rulers of Spain would have
seen themselves in diis light. But odier groupings are possible: if the French,
Italian and Spanish histories of this period appear highly regionalised and frag-
mented, German, English and eastern European histories appear much less so,
though the reasons are different in each case. German medievalists have been
litde troubled by ideas of revolution, feudal or otherwise; for them the decisive
break in European history comes in the second half of die eleventh century,
with the end of Ottoman and Salian rule, church reform, crusades and the
emergence of early scholasticism. Germany in the tenth century was as region-
alised as France or Italy or England, but the master narrative for its history is
still perceived as that of the history of kings. Although this has been rewritten
in the last generation with considerable sophistication and surprising detail, it is
still hardly linked at all to developments in social and economic history. 41 The
kinds of tenth-century developments which have impressed French, Italian
and Spanish medievalists - fortified aristocratic residences, the growth of
private jurisdiction, an increase in violence, the shift from slavery to serfdom -
can also be registered in the German long tenth century, but they are not seen
as having such significant consequences either for the course of events or for
the development of the polity.

Such conservatism should not be taken to mean stasis. A generation ago the
historiography of the German long tenth century did indeed not seem particu-
larly lively. The sources were both well edited and of known limitations, and it
was generally felt that, except perhaps for the ideology of rulership, where
there was evidently still mileage in continuing the lines of investigation opened
up by Schramm, Erdmann and Kantorowicz, there was little new to be said. If
today that no longer seems true, then this is not because of major discoveries
of source-material, or because the subject has received significant impulses
from outside: the debates on periodisation and revolutions have hardly
touched German historians at all. In retrospect, tire shift can be seen to have
been begun by Helmut Beumann’s study of Widukind of Corvey; 42 what this
triggered off over the next forty years was an increased sense of the need to

40 See note 39.

41 The largest recent survey, Fried (1 994), goes further in attempting such an approach than any previ-

ous survey; see also Fried (1991) It may be a sign of change that Fried’s neo-Lamprechtian approach

was not challenged, though other aspects of his work were: see Althoff (1995) and Fried (1995).

42 Beumann (1950).

Cambridge Histories Online © Cambridge University Press, 2008

TIMOTHY REUTER

read die great works of Ottoman historiography in dieir own terms. An almost
literary ‘close reading’ (though diis owed litde to literary scholarship and
nothing at all to post-structuralist views of the world, which have affected
German medievalists hardly at all) replaced what had become the increasingly
desperate interwar attempts to unpack these texts in a purely positivist manner,
to try to force them to reveal ‘how it really was’. At the same time, our under-
standing of the nuts and bolts of the east Frankish/ German kingdom was
transformed by detailed prosopographic investigations and by meticulous
reconstructions of die rulers’ itineraries . 43

England in the long tenth century was clearly as regionalised a society as any-
thing on the other side of the Channel. Indeed, it was in this period that
England came into being as anything more than an aspiration and perhaps on
occasions as a virtual community, and the process was not yet fully completed
by the early eleventh century . 44 Yet its historiography firmly resists a regionalis-
ing perspective; it is not that no such perspective has been offered, but rather
that there is no real place for it within the dominant discourse . 45 It might be
thought that the main reason for this is the sheer paucity of source-material:
the number of indisputably genuine tenth-century charters of all types from
the whole of Anglo-Saxon England hardly exceeds, die number of surviving
genuine diplomata issued by Otto I alone, and is a mere fraction of the number
surviving from the single if admittedly atypically rich archive of Cluny. The
richly symbolic accounts of east or west Frankish politics found in contempo-
rary narratives also have no surviving counterpart from Anglo-Saxon England.
More significant, though, is the influence of a dominant master-narrative, one
of English history as a success story made possible by the early development of
a strong centralising state. Recent historiography has fought hard to push back
the beginnings of this development beyond its traditional starting point in the
generations following the Norman Conquest, and a plausible case can be (and
has been) made for a ‘Carolingian’ phase of English history between Alfred
and Edgar, one in which military success, unification, legislation and the devel-
opment of what by early medieval standards was a fairly homogenous set of
local institutions went hand in hand . 46 Yet where an older generation of histo-
rians saw England as first dragged kicking and screaming into Europe, and
hence into modernity, as a result of the Norman Conquest, die new view has
rewritten tenth- and eleventh-century English history at one level whilst pre-
serving its isolation from continental developments at another. No kind of
mutation or revolution, feudal or otherwise, troubles the island, nor apparently

43 For the methodology and bibliography see Miiller-Mertens, chapter 9 below; see also Fleckenstein

(1966) and Leyser (1982b). 44 Wormald (1994).

45 For examples of regional studies see Stafford (1985), Gelling (1992), Yorke (1995).

46 See Campbell (1994), for the fullest recent statement of the view.

Cambridge Histories Online © Cambridge University Press, 2008

Introduction: reading the tenth century

do such things as the development of fortified residences or the freezing of
previously fluid settlement patterns, which remain by and large the concern of
archaeologists. 47

If the sources for English history in the long tenth century seem thin com-
pared with the wealth of the Mediterranean regions or even the plenty of tire
former Frankish kingdoms north of the Alps, they are rich compared with
those available for eastern and northern Europe. The histories of Rus' and of
the eastern European proto-states, ‘Poland’, ‘Bohemia’ and ‘Hungary’, are
probably the most contestable and contested of all those covered in this
volume. 48 This is partly the inevitable product of fragmentary information,
often late in date and highly ambiguous in its interpretation. But it is also, at
least for eastern Europe, a product of twentieth-century uncertainties. The
new states of the post-Versailles settlement have simply not enjoyed a continu-
ous existence over the last eighty years, unthreatened from without and con-
sensually accepted from within, and under such conditions it is not surprising
that historians of these regions have been slow to take up the methodological
novelties increasingly taken for granted further west. The histories of tenth-
century Poland, Hungary or Russia are as difficult to ‘read’ as those of sixth-
century Gaul or Britain — if anything, more so, since the written information
we have is almost all external as well as being late. But they are not so distant in
time and significance as are, for example, tire sixth-century Saxon kingdoms in
England; and interpretations of the fragmentary evidence are not as detached
from present-day reality and significance as they are for western European his-
torians, who inhabit societies whose sense of national identity does not require
a consensual view of a very distant past.

There remain the anomalous (from a western European perspective)
historiographical traditions of Byzantine history and European Islamic
history. 49 Though Byzantine history has a particular significance for Greeks
and Russians as the history of a ‘virtual precursor’, it is a more international
discipline than any of the areas of ‘national’ history so far studied. At tire same
time, the high demands it makes on its scholars’ linguistic and technical skills
have a double effect: few of its specialists have had tire time or energy to
become genuinely familiar with the history of western Europe (or even a part
of it) on the same level, while western medievalists have equally had to rely on
others as guides (as has the author of this chapter). None of the trajectories
which apply to the west really fit Byzantine history, for which the long tenth
century between 886 and 1025 is as much a golden age as an age of iron, in
recent interpretations not only politically and culturally, but also economically.

47 Hodges (1991) offers an outsider’s perspective on this.

48 For organisational reasons, the history of the Scandinavian lands was covered in NCMH n; see the

preface. 49 See chapters 22—5 (Jonathan Shepard) and 27 (Hugh Kennedy) below.

Cambridge Histories Online © Cambridge University Press, 2008

l6 TIMOTHY REUTER

Similar considerations apply to tire histories of the Spanish caliphate and the
Islamic amirs in Sicily in this period, except that here tire problem is com-
pounded by the fragmentary (and often late) nature of the source-material, and
by the politico-cultural significance of these regions, peripheries of a larger
culture whose metropolitan centre lay elsewhere. Nowhere in the area covered
by this volume is cross-cultural comparison more needed or more difficult to
carry out, from either side of the divide. In the present state of play, all that can
be said is that few of die periodisations and interpretative schemata which have
been applied to western Latin European history in the long tendi century
seem to have much relevance to Byzantine or Islamic history in the same
period, but that impression may neverdieless represent optical illusion rather
dian reality.

Some differences must have been real enough, however; the surviving
sources and traditions of interpretation no doubt exaggerate the extent to
which Byzantium (and its Bulgarian imitator) and Islamic Spain were societies
centred on a capital with a fixed court and a ruler who was much more than
primus inter pares, but no allowance one might make for diis could reduce diem
to the organisational status of the societies shaped by western European itiner-
ant rulership. Cultures which are urbanised and court-centred, whose rulers are
normally to be found at a fixed point from which they habitually tax and legis-
late, are inherently different from those of the main area covered in diis volume;
in particular, the antithesis of core and periphery (or of metropolis and prov-
ince) is a reality, not simply a metaphor.

The odier anomalous historiographical tradition is that of American medie-
valists (as it happens, hardly represented in the present volume, diough this is
the result of chance radier than calculation). Their traditions have not always
been clearly distinct from European ones; die first generations of American
medievalists were largely trained in and inspired by European schools of his-
torical writing, an intellectual dependency sustained in the mid-century era by
the influence of a number of important emigres and refugees, as elsewhere in
the American academy. But although the European medieval past is also
America’s medieval past, it is not its past in the same way. The links with
English history, and so, via the Anglo-Norman and Angevin empires, with
French medieval history have continued to be important, but diey are not the
only possible ways of appropriating die past. For Americans whose secondary
or primary ethnicity is eastern, central or southern European (there are very
few African-American or Asian-American medieval historians), they are not
even the most important ones. Moreover, the organisation of studies has
favoured a holistic approach to this particular past culture, taking in literary and
artistic remains as well as ‘straight’ history under the umbrella of Medieval
Studies, and in consequence exposing medieval historians to the influences of

Cambridge Histories Online © Cambridge University Press, 2008

Introduction: reading the tenth century

neighbouring disciplines in a way that is only beginning to happen in many
parts of Europe. Although American medievalists have taken sides in
European medievalists’ debates - and they have shown themselves just as liable
to Francocentrism as European historians - they have in many cases taken a
more detached and also a more innovative approach to the medieval past, and a
number of significant recent studies could probably only have been written
from the distance provided by the Adantic . 50

However fragmented the long tenth century may have been by the accidents
of source preservation and divergent historiographical traditions, there are still
generalisations which can be made about it, though, as we shall see, few are
uncontested. Estimates of changes in die level of economic activity in the long
tenth century have on the whole been moving upward in recent decades.
Monetisation is perceived positively; the Viking, Saracen and Magyar incursors
who caused Marc Bloch to depict the era in such gloomy terms are now
thought by many to have given positive impulses by raiding centres of accumu-
lated treasure and releasing it once more into economic circulation . 51
Population is also thought to have risen, though hard evidence is almost impos-
sible to come by. The beginnings of the urban renaissance which characterises
the high middle ages have also been sought in this period . 52 To the extent that
there is or can be any ‘pure’ economic history of this period, there is probably
more consensus about it at present than about any other aspect of the period.

Yet such developments are more easily described in a broad-brush sense
than explained. When we move on to social and political history in search of
explanations, consensus recedes. A number of other changes can apparently
be identified as characteristic of this period, and historians have been tempted
by die idea that many, perhaps even all of them can be linked in some way.
There is, first of all, the idea (Marxian in origin, though less so in its exposi-
tion or its specific application to the tenth century) that the long tenth century
saw a crucial shift away from slavery towards a serfdom which embraced not
only slaves but also a good part of what had previously been a free pea-
santry . 53 Second, we have the view, already mentioned, that settlement pat-
terns, previously fluid and shifting, solidified in this era. Linked with this we
have, third, the spread of the ‘private’, small-scale and residential fortification,
by contrast with die refuge fortifications of an earlier era, still being built
and planned in the late ninth century . 54 Fourth, such centres of aristocratic

50 Koziol (1992) and Geary (1994) are two examples; many more could be offered.

51 Duby (1 974), pp. 1 1 B-i 9.

52 Seejohanek, chapter 3 below, and also Hodges and Hobley (1988) and Verhulst (1993, 1994).

53 Bois (1989); Bonnassie (1991); see also, from rather different perspectives, Wickham (1984) and
Miiller-Mertens (1985).

54 Fossier (1982), pp. 182— 234; Toubert (1973a, 1973b); Bohme (1991a, 1991b).

Cambridge Histories Online © Cambridge University Press, 2008

TIMOTHY REUTER

domination were significant not only for the exercise of power but also for
shifts in family consciousness. Noble families defined themselves less in terms
of broad kindreds including relatives on both the male and die female side
and more in terms of a male descent lineage; these lineages often took their
names from the names of the fortifications which were the basis of their
power . 55 Fifth, the lordship exercised from these centres was often of a new
kind, based on pragmatic local dominance without much legitimation and cer-
tainly with little legitimation through ‘public’ office-holding. Rather, it came to
replace an older ‘public’ order which had survived in many regions from the
Carolingian era. This larger-scale public order was hollowed out to the point
of extinction in many parts of Europe during die long tenth century; royal
authority suffered earliest and worst, but it was followed into decline by the
authority of intermediate powers (dukes, counts, earls, archbishops,
bishops ). 56 Sixth, what remained was in essence ‘ties between man and man’:
legitimate authority had become privatised and personalised . 57 Linked with all
these developments was a seventh: the emergence of a new and enlarged
dominant class, a class which still had its own internal divisions but one in
which lords and their warrior followers increasingly perceived themselves as
members of a single group set apart from (and over) the rest of society; in the
course of the eleventh century a separate ideology and initiation rites would
be found for this class . 58

What all this adds up to is the totalising interpretation known as the ‘feudal
revolution’ or ‘feudal mutation’. It is a compelling view of the history of post-
Carolingian Europe (or at least of the history of Europe’s post-Carolingian
core); and yet for all its attractions it is a highly problematic one. Even leaving
aside those regions of northern and eastern Europe which were clearly follow-
ing another developmental trajectory altogether (as were Byzantium and Islam,
for quite different reasons), and in any case have not preserved the kind of evi-
dence which would enable us to form a judgement, the model does not really
seem to work for important parts of Europe: southern Italy, Leon, England,
Germany. As suggested above, this may be in part the product of different
historiographical traditions, though at least for England the model has been
explicidy rejected as inappropriate . 59 It is in any case a gross oversimplification
to call it ‘the model’: most historians working on this period would acknowl-
edge die existence of at least some of the phenomena enumerated in the previ-
ous paragraph and feel tempted by die idea that these phenomena were in
some way linked to one anodier, but, as already suggested, variations in empha-

55 Reuter (i 997a) provides a survey of the immense literature on this shift.

56 The essence of the ‘feudal mutation’; see Poly and Bournazel (1991).

57 The phrase was placed at the centre of interpretation, if not actually invented, by Bloch (1961).

58 Duby (1978); Flori (1979, 1983). 59 Campbell (1990).

Cambridge Histories Online © Cambridge University Press, 2008

Introduction: reading the tenth century 1 9

sis can produce considerable variations in die overarching interpretation which
provides the explanation of how these links actually worked.

Moreover, many of the most significant elements of the model are currendy
under challenge, even for the core regions of west Francia (including
Catalonia) and northern Italy, from which the model was derived. The chal-
lenges have intensified during the period between conception and publication
of the present volume. The extent of slavery in the early middle ages, and die
sense in which it was replaced in the long tenth century by serfdom, is highly
contentious. 60 So too is what once seemed common ground, the replacement
of public authority by personal ties, in other words ‘feudalism’. It has been
argued that feudalism, in the sense of a homogenous juridification of personal
relationships amongst the European governing elites, was an invention of the
twelfth century; fiefs and vassals, in this sense, were absent from die long tenth
century, and there was in any case no necessary link between vassalage and
benefice. 61 It is still not clear whether we should drink of a feudal revolution or
mutation at all; though Europe in 1100 was clearly very different from the
Europe of 800 or 900, not all would see the decades around the millennium as
marking a clear period in which most of the transition took place. 62 The con-
solidation of a small aristocracy and its warrior following into a single, wider
class was a process which does seem to have occurred across most of Europe
between the Carolingian era and the thirteenth century, but it was hardly a
homogenous or simultaneous one.

There are difficulties of perception here: are we dealing with new phenom-
ena, or merely with phenomena which began to be recorded more frequently
towards the end of the long tenth century? As local complaints of violence and
abuse increase, we are tempted to contrast them with an idealised Carolingian
past which may well never have existed, and which would appear quite different
to us were we to have as much information about its local look and feel as we
do about much of the post-Carolingian core of Europe around the millen-
nium. 63 Equally, die apparent fragmentation of large-scale political authority in
many parts of Europe may indicate a new order, but at least at the regional level
the polities of this period (notably the French, German and Italian principal-
ities) were in most cases not arbitrary creations but had much older roots as
vehicles of being and consciousness, often traceable back through the
Carolingian era to the early middle ages. It is even conceivable that the smaller
units of lordship which become clearly visible around the millennium had

60 Verhulst (1991); Barthelemy (1995); see also the symposium of responses to Bois (1989) in Medievales

21 (1991). 61 Reynolds (1994); for initial responses see Nortier (1996); Barthelemy (1997).

62 Bisson (1994), with responses by White (1996), Barthelemy (1996), Reuter (1997b), Wickham (1997)

and a reply by Bisson (1997) See also the exchange between Barthelemy (1992a) and Poly and

Bournazel (1994). 63 White (1996), pp. 218-23; Reuter (1997b), pp. 178—87.

Cambridge Histories Online © Cambridge University Press, 2008

TIMOTHY REUTER

older roots, now lost to sight. Any attempt to offer a synthesis at this stage
would be futile; as historians from different European traditions become more
aware than they have been of each other’s practices and findings, and as inter-
est in the period around the first millennium finds at least a temporary increase
in public and professional awareness from our contemplation of our own posi-
tion in the decades around the second millennium, debate on these issues,
which are also of central importance to historians of the periods preceding
and following the long tenth century, is likely to intensify and to shift as it does
so.

If we leave die awkward terrain of social and political history and turn to
religious history, then we might at first think that the history of the church in
this period would appear to be a good example of encellukment , at least at a
purely institutional level. Ninth-century popes had commanded and occasion-
ally threatened bishops; they had deposed or confirmed some of them in
office; at least a few had been significant figures who could not easily be
bypassed. But papal leadership of Christianity was far more muted in the
period which followed. Ecclesiastics might journey to Rome on pilgrimage, but
they mostly setded their own affairs. Neidier the existence of a papal judge-
ment nor the presence of a papal legate was necessarily bankable capital in the
course of a dispute, and the privileges granted by popes were more than once
in this period publicly repudiated. This was not so much a rejection of the pope
qua pope as a refiection of a more general attitude which meant that the
members of die higher ranks of the church hierarchy were largely insignificant
except in their capacity as bishops. Councils were rare, and usually local affairs
when they did meet: bishops were largely sovereign widiin their own dioceses,
and were the crucial figures of the tenth-century church, as Rosamond
McKitterick’s chapter demonstrates.

Ecclesiastical encellukment was also visible, in a sense, in die history of mon-
asticism in this period. Historians have been able to free themselves, slowly,
from the notion that monastic reform in this period was spelled Cluny; but it
has been more difficult to dispel ideas of monastic ‘orders’ projected back
from the twelfth century and later. Yet even Cluny’s collection of monasteries
widi varying ties of dependence on it was not an order in the later sense: the
ordo Cluniacensis was, as Joachim Wollasch points out, Cluny’s ‘way of life’, not a
legally defined body. Odier monastic groupings were still less institutionalised,
depending as they generally did on die attentions of a reforming ‘expert’. Yet
the very existence of such ‘experts’, men like Gerard of Brogne or William of
Volpiano, shows how encellukment was not all-determining. Even if such
monastic families had a short-lived and tenuous existence, they could link and
unite, however briefly, monasteries scattered over several dioceses, even king-
doms. The elite owners of monasteries, especially when these were bishops,

Cambridge Histories Online © Cambridge University Press, 2008

Introduction: reading the tenth century 2 i

were sufficiently knowledgeable to be able to see beyond the rim of their
immediate locality.

The localism which is such an evident part of church life in this period was
not transcended in the field of monastic life alone. Historians have been
inclined to see it as a demonstration of how entangled with the affairs of the
world the church became in the post-Carolingian era and certainly there is
enough anecdotal evidence of abuse and gross misconduct, at least from some
regions of Europe, to support such a view. Yet it is also possible to read the
history of tenth-century Christianity as one of remarkable success. 64 It was not
only the era in which die Carolingians’ attempts to convert the regions beyond
the former territory of die Roman empire were continued and largely brought
to completion, but also the one in which the Christianisation of Europe’s inter-
ior finally became reality. It is not so much the evidence from the period which
follows that demonstrates diis: evidence for an insistence by enlightened laity
and clergy alike on the ‘Gregorian’ diemes of a sexually pure clergy and a
church untainted by die moral corrosion of payments in cash or in favours. It is
also die emergence of a more active lay participation in Christianity, which
took many forms: large-scale and long-distance pilgrimages, notably to Rome
and Jerusalem; the veneration of relics on a very substantial scale; arguably also
the mass participation in die movements known as die ‘Peace’ and ‘Truce of
God’, though this is stressed much more by some contemporary observers,
notably by Adhemar of Chabannes and Radulf Glaber, than by others. 65 Even
some of the heresies of the period (and the recording of heresy from about
1000 onwards is itself a novelty) are interpretable in terms of ‘leftist deviation’,
as the products of people who have been reached so effectively by the message
as to take it too far; the same is true of die occasional notes of anti-semitism of
the period. There is a note of questioning, of self-doubt, in the writings of
many ecclesiastics of diis period - Radier of Verona, Thietmar of Merseburg,
Wulfstan of York — which seems both more strident and more searching than it
had been in the Carolingian era. And aldiough it is clear that many of those
who lived around die eschatologically significant dates of 1000 and 1033 did
not do so in fear (or hope) of die Second Coming, it is, at the end of the second
millennium, less clear than it seemed to Ferdinand Lot and his contemporaries
that no one at all did. It is more likely that the intensification of religious expe-
rience around die millennium, perceptible in a number of ways, was, at least in
part, a response to the millennium itself. 66

Culturally and intellectually the period has often been seen as one of

64 As stressed by Tellenbach (1993).

65 Head and Landes (1993) stress the links between the peace movement and other aspects of popular

religiosity; see also Moore (1980) and Leyser (1994c).

66 Landes (1988), Fried (1989), Landes (1992, 1993, 1 995).

Cambridge Histories Online © Cambridge University Press, 2008

TIMOTHY REUTER

stagnation, even though it was also a period in which some members of the
elite invested massively in the arts, whether in buildings, ivorywork, goldsmith-
ery and other metalwork, or illuminated manuscripts, as Henry Mayr-Harting
demonstrates. The notion of intellectual stagnation rests on rather superficial
judgements: some of the yardsticks which have been used, such as die copying
of manuscripts, are inappropriate, and in any case there was, at the level of elite
culture, more happening than at first meets the eye, as Claudio Leonardi dem-
onstrates. Nevertheless, there are some signs of decline which are difficult to
deny, the most important of which is the reduced importance of schools
across Latin-speaking Europe (the trajectories of Byzantine and Islamic intel-
lectual history are not covered here). There were fewer schools, so far as we can
judge, than there had been in the ninth century; still more significant, their con-
tinuous existence was increasingly precarious and fragile, dependent on the
isolated, often highly charismatic figures who had built them up and whom
they rarely if ever outlived. This fragmentation and impermanence may
perhaps be taken as a cultural and intellectual mirroring of encellulement. So also
may the decline in the importance of courts as centres of cultural and intellec-
tual production. It is true that modern historians have tended to use the term
court as a shorthand for a set of activities in some way connected and intercon-
nected by a ruler and his entourage, thus making courts (like scriptoria) as
much a modern social construct as a Carolingian reality. Yet, even making
allowance for die gap between present construct and past reality, it remains
evident that tire royal and princely entourages of the long tenth century had
given up much of the functionality of their Carolingian predecessors.

The period has also been seen as one with a sharp decline in pragmatic liter-
acy and a consequent increase in the importance of symbolic and non-verbal
forms of communication, though this is a problematic view for two reasons.
The decline in pragmatic literacy was in regional terms a very uneven affair . 6 It
is not evident that there was much decline, if any, in Italy, or Spain, or
Mediterranean France. The paucity of source-material for Anglo-Saxon
England is more likely to be the product of post-Conquest neglect of and con-
tempt for the Anglo-Saxon past, which increasingly lacked any legal signifi-
cance, than of any lack of production at the time. Indeed, it is clear both from
contemporary indirect evidence and from later fragments and fossilised prac-
tices that tenth-century England must have made extensive use of the written
word . 68 Since the newly converted lands on the northern and eastern peripher-
ies had, for all practical purposes, not known literacy previously, the downward
curve in the graph of pragmatic literacy really only describes the position in the

67 The contributions to McKitterick (1991) provide the best survey.

68 Wormald (1977); Kelly (1991); Keynes (1991).

Cambridge Histories Online © Cambridge University Press, 2008

Introduction: reading the tenth century

2 3

former Frankish kingdoms north of die Alps, and even here it applies mainly
to the first half or two-thirds of the century.

As to the use of symbolic and non-verbal forms of communication, diese
were indeed important in the long tenth century. But, so far as we can judge,
they were just as important in those regions which continued to make extensive
use of writing. Moreover, they were also important in the periods which pre-
ceded and followed diem. It is tempting to see the period as one in which poli-
tics found expression through liturgy rather than law, or as one dominated by
ritual, ceremony and gesture 69 but it would be more accurate to say that histori-
ans’ eyes for such tilings have been sharpened in a period superficially poor in
other kinds of source, whereas their presence has been more readily over-
looked in seemingly more articulate eras like the ninth or twelfth centuries.
Whether in the ninth, the tenth or die twelfth centuries, the primary function
of social and political ritual was in any case not to act as a substitute for writing
as such, but rather to make actions visible and permanent to non-literate lay
elites whose members had no other means of defining them and fixing them in
memory. Nevertheless, such a cultural approach, rethinking political history
through a study of the seemingly inconsequential details of ritualised behavi-
our, has been of particular significance for the long tenth century for historio-
graphical reasons: however wide the potential applicability of the technique, it
happens to have been tested most thoroughly on this period . 70

How, then, should we ‘read’ the tenth century? If it is indeed possible to
make out at least some trends with a general significance across the period and
region, does that mean that we can and should resume the search for general
interpretations? The first point is that it is possible to read die period. The
‘obscure’ or ‘dark’ age is less dark than it seems, in spite of the shortage of
large-scale contemporary narratives to provide an initial interpretation (or
rather, of rival narratives to provide alternative interpretations). The
Braudelian longue duree and the medium-term flow are well-enough docu-
mented; it is the surface play of the political which is frequently less well
recorded. The difficulties lie - to continue the metaphor of reading - not so
much in die aporias, in the letters, words, even whole sentences and paragraphs
which are missing from the ‘text’, but rather that we are not always certain what
the pieces of ‘text’ which survive really mean. Literal readings, in other words,
are frequently either not possible or merely sterile. Much of what has been pre-
served from the tenth century simply will not yield to a common-sense under-
standing, and this is true of the apparently straightforward as well as of the
evidently obscure or non-literal.

But reading is difficult, nevertheless, because few of the generalisations

69 For the first interpretation see Kantorowicz (1957), pp. 87—93; for the second, Leyser (i994d).

70 Althoff (1990); Koziol (1992); Leyser (i994d); Althoff (1997).

Cambridge Histories Online © Cambridge University Press, 2008

TIMOTHY REUTER

which have been offered, at any level, seem to work for all of Europe, even
after we have made allowance for the distorting effects of national historio-
graphical traditions. In that sense, encellukment'mcs, a reality: we are dealing with
regions which did not necessarily transmit their developments to their neigh-
bours, or receive and absorb their neighbours’ developments - or not with any
speed. And yet, at the level of elite culture at least, the post-Carolingian core of
Europe showed a remarkable degree of homogeneity and internationality.
There are clear regional flavours to such material remains as writing or building
styles, yet they are evidently precisely that: regional flavours, not autonomous
practices. It was this post-Carolingian core also which from this period on
would provide die model adopted by the emergent societies of northern and
eastern Europe ; 71 at this time diey were still locked in an encellulement far deeper
than anything found in the west, even if this is now concealed from us by the
homogenising effects of ignorance. And it was this post-Carolingian core
which came to define itself through opposition to the older, rival
Mediterranean cultures of Byzantium and Islam. It would not, as yet, contest
their dominance or do more than nibble at their territorial edges, but die sense
of difference, so visible from the later eleventh century onwards, was already
beginning to form in the period covered in this volume.

Whatever level or form of European history we examine in this period, we
appear to be confronted by past behaviour which presents itself at once as
having been highly unsystematic and locally specific and as having been wide-
spread: it is this paradoxical relationship between coherence and fragmentation
which in die last resort dominates almost all readings of the long tendi century.
The reader of the chapters which follow will do well to bear this paradox in
mind, and will also do well, in approaching this collective reading of the tenth
century, to think of the period not as ‘pre’ or ‘post’ anything, but radier as of
itself. It is difficult enough, and rewarding enough, even when approached on
these terms.

71 Bartlett (1993).

Cambridge Histories Online © Cambridge University Press, 2008

PART I

GENERAL THEMES

Cambridge Histories Online © Cambridge University Press, 2008

CHAPTER 2

RURAL ECONOMY AND
COUNTRY LIFE

Robert Fossier

even though current historiography still preserves the concepts of ancient
and medieval history devised in the nineteenth century, it is becoming steadily
more apparent that these divisions - generally drawn up with political history
in mind — are unsatisfactory for the historian of the economy or society. In
these fields there is a longue duree from the decline of slavery in the third century
to the first significant use of machine power in the eighteenth century.
Nevertheless, undeniable developments in the techniques of production or in
the relations between men force us to mark out certain stages in this long
period, one in which Europe entered on the stage of world history. At what
point was there a transition from the shrunken and undynamic structures still
associated with Germanic or Graeco-Roman custom (the two were in this
respect very similar) to structures in which the relationships between men and
men and a production generating profits announced a more ‘modern’ eco-
nomic climate? The question is not otiose; the answer will determine the view
one takes on the ‘infancy of Europe’. In fact, almost all the observations which
one can make, whatever the preoccupations of individual historians, point to
the tenth century as the age of growth, of take-off, of rising, or some such
phrase. In 898 we find the word feodum used in southern France to mean a
tenancy by military service; in 910 the foundation of Cluny opened up a new
phase in die history of spirituality; in 920 villages began to move on to hilltops
in central Italy; in 95 5 the Magyars were definitively beaten; in 970 the series of
commercial contracts surviving from Venice began; in 980 the gold of the
Catalan parias arrived at Barcelona, and there are other similar examples from
all spheres of economic activity.

This transformation of die old world was indeed a ‘revolution’, if one is pre-
pared to concede that the word does not have the same implications as it does
in our own epoch but refers to a slow, indeed a very slow, transformation of the
framework of human life. The judgements made by historians on this major
turning point in the history of Europe are often marked by scholars’ own

2 7

Cambridge Histories Online © Cambridge University Press, 2008

ROBERT FOSSIER

philosophical convictions. Those persuaded of die fundamental correctness
of Marx’s analysis will see here the beginnings of a ‘feudal’ era, which set up,
often violently, a kind of tacit contract between a lord who protects and a
worker who feeds him; others who remain faithful to ‘Romanist’ theories will
see a generally peaceful transformation from the structures inherited from
antiquity to a newer version, determined by conditions equally new; others
again will refuse to believe in this transformation and search ardendy for
proofs of continuity. I find it hard to believe that this latter group can be in the
right: it seems to me fairly evident that a new order did indeed establish itself,
one which did so with all the slowness familiar to the historical anthropologist
but which nevertheless gradually coloured a society nine-tenths of which, it
should hardly be necessary to repeat, lived in die countryside. It is necessary to
begin by saying a few words about these country-dwellers.

First of all, the two feelings which until then had oppressed everyone in
Europe - fear and violence - still dominated. Enthusiasm was not on the
agenda; it may be that by the thirteenth century both feelings had come to be
held in check, but it is hardly credible to say this about the millennium, even if
the well-known ‘Terrors’ of that fateful year were produced by the dreams of
romantic historians. Shortages constantly threatened; one could even say that
with the population growing faster than technical progress, their grip tight-
ened; die acts of cannibalism noted for 103 3 are a well-known example of this.
The fear of want, that fear which prostrated the faithful before an oppressive
and vengeful God, did not end, then. Nonetheless, some solutions began to
appear in relationships of neighbourhood, profession and family, which we
shall return to later. As for die violence of the armati (warriors) and the ‘terror-
ism’ which they have been described as exercising: the barriers erected against
it - justice and the Peace of God - were as yet perhaps not very effective, but
unbridled vendetta and constant plunderings were on the decline after 1050 or
so. The raiding by warriors, mrra as it is known in the texts, continued to wreak
havoc and misfortune, but it tended towards feud rather than ‘warfare’.
Gratuitous cruelty and sadism were becoming individual rather than collective
failings. The superiority of the stronger expressed itself increasingly in repre-
sentative and symbolic behaviour: one had to astonish and provoke the admi-
ration of those who could no longer be exploited unrestrictedly. To eat more
than necessary, to distribute alms and gifts, not to move about except with a
vast entourage, these were the marks of the dominant, die ‘noble’ man. In such
a world of gift-giving, well known to anthropologists, the gesture took on its
full value as a symbol: it validated all serious commitment, replaced writing,
which was only just beginning to revive, and even speech; the latter, even in the
form of oaths, only acquired force from the gestures which accompanied it. A
final point, perhaps the most important one here: in southern Europe there

Cambridge Histories Online © Cambridge University Press, 2008

Rural economy and country life

2 9

was a written law, whether or not we call it Roman, and in the north such law
had also been introduced in the writing of law-codes. But who knew how to
read these except for a handful of professionals? It follows that the lot of men
was largely governed by custom, both spoken and performed. Day-to-day atti-
tudes were shaped by the past, all novelty being in principle both bad and dan-
gerous; this conservatism of spirit was appropriate to a society slow to move.
Historians may well attempt to classify individuals into small juridically defined
groups, but in fact, in this period, people were what other people thought they
were.

The slow expansion of this period, which was only just beginning before the
millennium, presents historians, however aware they may be of the issue, with
two problems which have still not been fully resolved. First, they are tempted
to place the beginnings of these developments around die middle of the tenth
century, drat hidden turning point of medieval history. Here sources are so thin
on the ground, especially north of die Loire and of die Alps, that one has to
say in all honesty that we can assume but cannot prove. For diis reason there is
no shortage of historians to discuss the role played by the Carolingian era.
Contrary to what is believed by German historians in particular, the role of die
Carolingian dynasty is not a particularly interesting topic: its effects beyond the
Channel or the Pyrenees, and even in southern France or Italy, were non-exis-
tent or negligible. But it is worth talking about the importance one should give
to die period between 700 and 850 as a harbinger of things to come. Anglo-
Saxon kings, Frisian merchants, Iberian princes, the Frankish aristocracy, the
litterati of Italy created a movement which pre-dated Charlemagne, and one
might even say that they made him possible. In matters of canon law, in die
reinforcement of the nuclear family, in the reform of the church, in the revival
of the role of the state, in the taste for antique culture, die period was not neg-
ligible. Neverdieless, I do not believe that die changes brought about were uni-
versal, and in the two areas to be dealt with here, the economy and rural society,
the legacy of the ‘Carolingian era’ was minimal, and litde of it can be traced
after about 9 50 or 1000. A ‘mere surface ripple’, as Georges Duby has put it. 1

The odier problem is probably still more difficult: that of the ‘causes’ of the
European awakening. This is a classic demonstration of die chicken-and-egg
problem: what was cause, and what effect? Technical progress? But how can we
determine it? An easing off of the assaults which had afflicted Europe since die
third century? But there were still Vikings or Normans in die tenth and
eleventh centuries, Magyars until the millennium and Saracens until the end of
the eleventh century, quite apart from the internal iverra, which was hardly a
peaceful affair. Demographic expansion is something more certain, and we

1 Duby ( 1 973), p. 121 (— Duby (1974), p. 106).

Cambridge Histories Online © Cambridge University Press, 2008

3 °

ROBERT FOSSIER

shall return to it, but where, when and why? Perhaps we can talk of a slight
improvement in the climatic conditions in western Europe, favouring an
increase in plant life and in human and animal life as well, because this is a
datum which is clear, unquestionable and certainly not without effect. The evi-
dence is indisputable: from 850/ 900 onwards the beech climbs the foothills of
the Alps and the Bohemian mountains; the birch yields ground in Scotland and
Scandinavia; both the sea and malaria retreat in the coastal marshes. I am not
competent to say why these changes took place, but the Christians of this
period, if they noticed these phenomena, might perhaps have seen in them the
sign of a God finally appeased.

These Christians in their turn, even if they have not left us very numerous
sources, did indeed note some of the essential changes which struck them:
demographic growth, new family structures and the establishment of fortified
residences are referred to in hagiography, biography and historiography, while
charters register changes of fortune or status. Iconography remained impover-
ished and conventionalised, but rural archaeology has compensated for this,
and in the last fifty years, especially in northern and north-western Europe, has
provided new evidence on human habitat, tools and utensils. The ‘dark ages’
are lightening a little.

HUMAN GROUPS

Small groups of people, not very numerous, grouped around a paterfamilias , a
clan chief or a lord, separated by huge areas which were not or scarcely
exploited, short of tools and especially of iron ones, scratching a meagre living
from the soil with difficulty: that is the countryside of die early middle ages.
Here and there, there were larger-scale lordly estates, die descendants of the
villae of the Romanised regions or of the curtes of the barbarian era, worked by
slaves. Blinded by classical towns or fascinated by the mosaics which decorated
a few exceptionally rich houses, ancient historians have refused to admit, in
spite of archaeological evidence, that to die left and to the right of the Rhine,
the Germanic and die Graeco-Roman rural economy were in much the same
state. In what follows we shall see how diese somewhat unpromising character-
istics were to soften and reform.

Constraints and relaxations within the family

The basic unit of daily life was the family. Prior both to the state, which it
ignored, and to the parish, which was just in the process of formation, it repre-
sented the basic unit of production because there was still no exterior element
which could replace the labour force of the family group within the rural

Cambridge Histories Online © Cambridge University Press, 2008

Rural economy and country life 31

economy. The history of this group has made substantial progress since 1980:
careful study of prosopographical evidence or of the genealogies drawn up in
the eleventh and twelfth centuries allows us to identify the various forms of
kinship structure which succeeded one another. Even archaeology has made
some contribution by dating the settlement forms of the period. The term
familia, used conventionally by all scribes in dais period, is maddeningly vague:
the group it denoted might be limited to the nuclear family or extend to very
distant clients, and it is probably better not to pursue the matter here.

The historian who examines the family of die first half of the tenth century
will find three levels. The ‘clan’ (German Sippe) includes all individuals of the
same blood in the sense that they recognised a common ancestor: this might be
a group of a thousand or more men and women most of whom had no contin-
uous daily relationships with each other. Sometimes a particular group - we
may call it a tribe - imposed its authority on others, as for example when
seeking new lands to pass dirough or cultivate. This pattern was the most
archaic, and was found at a very primitive economic level, that of peripatetic
hunters and gatherers. Such structures have left traces in the epic literature of
the early middle ages; by 900 they were scarcely to be found outside Scodand,
Frisia, Scandinavia, and possibly Brittany. Such systems were essentially cog-
natic, even though women could be excluded from any public role.

The setded clan breaks up into lineages (Latin gens, German Geschlechfy. here
a real or mydiical but not too distant ancestor defines the group of blood-rela-
tives. Here we are at the heart of the family structure which took shape in the
early middle ages in the Christian west. The lineage was sedentary, endoga-
mous and conscious of the purity of its blood, and lived in compact groups of
perhaps a hundred members, as the occasional excavation of a great hall of the
seventh or eighth century shows. These families were warriors or peasants,
hence male-dominated, but because women were evidendy the guarantors of
purity of lineage their position was better, though they were subject to close
surveillance. One can see that this family type remained that of the dominant
class, for it was precisely in diese kin-groups that they found the elements they
needed to maintain dieir domination over others.

When at the end of the eleventh or in the twelfth century many ‘noble’ fam-
ilies wanted to establish their genealogy, memory did not take them back
beyond a ‘wall’ in the past - 800-50 for the very noble, 950-80 for the less
noble - beyond which it was necessary to invent. This is the result of a second
significant shift in family structures, in the late nindi and tenth centuries. The
lineage in turn crumbles, breaks up and forgets itself. We are now confronted
by die ‘house’ ( Haus , maison, domus), with a simple direct line of descent (even
when parallel lines also exist). Everything suggests that this simple structure,
based on the nuclear family, formed gradually within the lineage, and that if the

Cambridge Histories Online © Cambridge University Press, 2008

3 2

ROBERT FOSSIER

latter disappeared before die millennium it was as a result of this internal sub-
division. The more humble members of the population, to whom we shall
return, did not share this preoccupation with the purity of descent. The conju-
gal system was not simply supported by the teachings of die church; it corre-
sponded to the position of die vast majority, and its triumph would soon come.

It will be recognised that so profound a set of changes did not come about
at a stroke in die period with which we are dealing. One of the most visible
characteristics of family life around the millennium, and one which would
continue to be important long after diat, was the powerful constraints exer-
cised by the kindred, not only within noble families, where group interests
checked ill-considered personal initiatives, but also among the most humble
(at least among those to whom our documents bear witness - a significant
restriction). Such constraints could take anodyne forms such as the passing on
of die same name from generation to generation, the daily bread of prosopog-
raphers in search of lines of descent. Or they could acquire moral or Christian
dimensions, which restricted certain freedoms, for example the freedom to
bury: one had to lie where one’s ancestors were gathered. The earliest exam-
ples of genealogical literature - for the counts of Flanders in the mid-tenth
century, for the lords of Vendome at the end of the tendi century - stress this
link with one’s ancestors. A count of Anjou in the twelfth century was to say:
‘Before this I know nothing, for I do not know where my forefadiers are
buried.’ 2 A final superficial point: it is from the end of the tenth century that
we begin to find signs in dress and emblems distinguishing one ‘house’ from
another. But it is obvious that it was in the economic sphere that the power of
kindred was most important, all the more because the contemporaray shifts in
landholding, to which we shall return shortly, threatened to shake the base of
its wealth and hence its power. The laudatio parentum, by which descendants or
collateral relatives gave their consent to property transactions or gifts of land
in alms, offers some rough figures, even granted that what we have is only the
positive side (should the kin refuse their consent, the transaction would not
take place and diere would thus be no record of it). In Latium, the proportion
of transactions with a collective nature or mentioning the approval of kindred
was 35% between 900 and 950 and 46% at the end of the tenth century, and
still stood at 41% around 1050. In Catalonia, shordy before die millennium,
die figure was a mere 1 2%, but rose to 30% fifty years later. In Picardy, where
fewer documents have survived, it rose from 17% to 36% in the same period.
Clearly there are gaps and variations, but the role played by the lineage
remained significant, perhaps even grew. Epic literature is replete with exam-
ples, from the four inseparable sons of Aymon, to the family of Ganelon, all

2 Fulk le Rechin, in Chronique des comtes d Anjou et des seigneurs d’A.mboise , ed. L. Halphen and R.

Poupardin, p. 237 .

Cambridge Histories Online © Cambridge University Press, 2008

Rural economy and country life 33

executed along with the traitor, though it is true that this evidence comes from
a rather later period.

However, further constraints began to take effect; they were already in place,
even though they did not play much of a role before 1080 or 1100.
Primogeniture was one, especially for the richer members of society; the lower
orders did not here have the motivation driving the dominant classes. The aris-
tocracy could not henceforth with equanimity run die risk of their heirs’ joint
partition on their death, a source of innumerable difficulties, or of a division of
estates. The designation of a preferred heir was not new: Roman testamentary
law allowed it. But die combination of an emphasis on male succession and
primogeniture was destined to preserve the integrity of family possessions
against younger sons and kinsfolk: diere are examples from die Loire valley in
the 1010s and the practice would spread across north-western Europe in the
first instance during the following century. A furdier path was opened up by
the freeing of marriage. By taking wives exogamously, outside one’s own kin-
group, and by founding the union not on kin interest but on mutual attraction
( dikctio ) and free will ( consensus ), one might hope to escape the intervention of
relatives. Such a practice was obviously of interest for the humblest; Roman
law, moreover, as well as die Bible, enjoined it. Carolingian church legislation
had already laid down exogamy as the norm, and in 1025 Gerard of Cambrai
was to back this up with sanctions. 3 It was a key force for the liberation of the
individual. But the pressure exerted by the lineage mentioned earlier shows
that it would take a long time to triumph.

THE CONTINUING STRENGTH OF FREEDOM

The church asserted it, and both Roman and Germanic law testified to it: the
‘normal’ man was a free man. He had free disposition of his own body, his
goods, his arms; he assisted in die doing of justice, and sustained the prince.
Even if he held lands by service, he could leave them and take up another
holding. The ideal culminated in the hermit, released from all control and
perhaps for that very reason venerated and consulted, as Otto III did St
Romuald. Nevertheless, once military skill demanded time for training which
work on the fields did not allow, once the complexity of cases and laws coming
before tribunals could only be determined by experts, and once the difficulties
of subsistence farming meant that workers could not take time off whenever
they felt like it — in short, once liberty found its limits — the Roman notion of
freedom would become no more than a word. And this was indeed the position
for a very large number of people in the tenth century, though not without
variations.

3 Acta synodi Atrebatensi a Gerardo Cameracensi etAtrebatensi episcopo celebrata anno 102J, pp. 2—5.

Cambridge Histories Online © Cambridge University Press, 2008

ROBERT FOSSIER

Leaving aside for die moment those whose wealth or office placed them
beyond all constraints and enabled them to command others, die group of free
men, the overwhelming majority of the population, can be divided into at least
three groups, which can be distinguished juridically more easily than economi-
cally. The highest group was made up of those who were subject only to public
authority, either of the count or of die king, who presided at tribunals and gave
their opinion as well as enjoying established rights over public land: the allodial
landowners (from al-od, property held absolutely). They will be known as
Schoffenbarfreien at the Germanic mallus, sokmen at the Anglo-Saxon shiregemot, ari-
manni in the Lombard plain, boni homines in Catalonia, or simply liberi, as in Gaul.
Since they did not make extensive donations to the church their traces in the
documents surviving to us are limited; but the loci in northern France which are
not part of large estates, the contracts of thirty years’ rent followed by outright
ownership which formed the aprisio of Languedoc, die pressura of Catalonia,
and the escalio of Aragon (this whole region is rich in documentation) from
about 820 or 850 onwards, the isolated casalia of Tuscany, Sabina and die banks
of the Po, as well as the casae of the Auvergne and perhaps the Breton ran, all
show the vitality of die free peasantry. Archaeology has demonstrated, in the
proto-villages of the ninth and tenth centuries, the existence of large family
enclosures which were certainly not dependent on a lord. In turn, the libri tradi-
tionum of northern France or the Empire beyond the Rhine multiply examples
of the rights held by these villani or pagenses-. access to common lands or to
assarts which were to encroach on die commons at the end of the tenth
century, in Cerdana, Normandy and northern Italy. One would like to be able
to assess the proportion which this large group represented: as has been said,
the sources are often silent, but in Catalonia around 990 (80%) of the charters
recording exchanges in favour of the church concern it, and fifty years later in
the Maconnais the figure would be 60%. At the point when seigneurial organ-
isation was being established, this group clearly presented an obstacle, and
there are indeed numerous signs of the efforts of the powerful to force the
allodial peasantry to submit themselves to their authority, either by personal
commendation or by ‘receiving’ their lands back in dependent tenure. We find
these from 975 in Provence, from 1010 in Tuscany and the Thames valley, and
from 1040 in Latium and the Loire valley. Public authority, however, had an
interest in supporting such free men, and in those regions where comital power
remained strong that is what counts did: in the valleys of the Scheldt, Meuse
and Rhine, in Bavaria and Saxony, allodial lands {Eigen) remained protected.

Nevertheless, the importance of large estates, to which we shall return, is
undeniable. Even if he made use of some domestic labour, the master was
forced to lease out lands. This was a very ancient ‘system’, known from the
times of antiquity with its peasant coloni who paid a rent {canon or tasca ) on the

Cambridge Histories Online © Cambridge University Press, 2008

Rural economy and country life

holdings granted to diem and who came to owe - but from when? - labour ser-
vices on their proprietor’s own lands. Since we know of die status of diese
men, for the most part juridically free but all economically dependent, only
from the texts of the ninth century, especially die great (so-called ‘Carolingian’)
polyptichs, there have been and still are fierce scholarly debates about them,
which can be passed over here, since from 900 or 950 at the latest the silence of
the sources is so profound that the debate is irrelevant to us. What we appear to
see, around the millennium, is fairly simple. Either in nuclear families or in
groups, people held, in return for rent, tenancies thought capable of sustaining
them and yielding a surplus with which they could meet dieir obligations to
their lords: the Germanic hob a, die Anglo-Saxon hide, the mansus of Romance-
speaking lands or the Italian colonica. There is no space here to go into the prob-
lems of surface area (ranging from 2 to 24 hectares!), of service obligations, of
the nature of renders or the right to commute services. Some of these prob-
lems are connected with the origins of the seigneurie, and will be dealt with
there. We shall also not spend much time on certain kinds of tenancy, as for
example those used by the church (the censuales or sainteurs of Lotharingia and
the lands beyond the Rhine), which the church alone thought were more
favourable than odier kinds, because in theory their holders were covered by
the mantle of ecclesiastical protection, though the price for this was heavy per-
sonal taxation. The common characteristic of all these men was straightfor-
ward: tliey were free, served in die army, and were perhaps still able to make
themselves heard in public judicial assemblies. But they were weighed down by
severe financial exactions which set severe limits on their freedom: they could
leave, they could choose wives from elsewhere, but they were then excluded
from the group which surrounded them. Their freedom, it has been said, was
‘the freedom to choose their lord’. 4

Some became irrevocably attached to their lords and so came to form a third
group: these were men of service (domestics, Dienstmdnner ) charged by their
lords with tasks of administration or supervision ( officiates , ministeriaks).
Generally free (though in Germany they were to become servile), their
numbers expanded once the relaxation of the ‘system’ forced their masters to
delegate tasks. One finds them from between 920 and 1000, throughout the
zone stretching from Burgundy to Bavaria and from the Alps to the lower
Rhine, at the head of outlying estates, as ‘mayors’ or supervisors; some also had
military careers. Normally they lived from a portion of land granted by their
lord and from dues, which makes us suppose that they often abused their posi-
tions and were detested.

Allodial landowners, tenants, officials: these men were free, but free in the

4 A phrase used by the Marxist historian Bessmertniy (i 976 ).

Cambridge Histories Online © Cambridge University Press, 2008

ROBERT FOSSIER

shadow of die powerful. This is why diey were tempted to gadier round the
powerful and support them. We are still a long way from the rural assemblies of
the twelfth century or even from the active confraternities of the eleventh, but
the seeds for these developments were sown in the tenth century, in various
ways. First, some of these groups were able to obtain tenancy arrangements
which were particularly supple and advantageous. This was the case with a
number of southern groups: the libellarii of northern Italy or Umbria at the end
of the tenth century, and the aprisionarii of Catalonia and Languedoc fifty years
earlier. Once the emphyteutic lease was over, they could become owners of a
part of their holdings, and enjoyed guarantees of justice and free access to the
saltus , communia , terra francorum, even freedom from mercantile tolls. Further
north the situation was less favourable, but it is there that we can find convivial
meetings providing the basis for conscious solidarity. Groups formed around a
patron saint or a drinking-bout are clearly visible from the tenth century
onwards at London, Exeter and Cambridge, and scarcely later on the
Scandinavian coasts (Birka and Hedeby), then down the Rhine and at St
Gallen, and in the early eleventh century in die Low Countries, in the valleys of
the Meuse and the Scheldt. True, these drykkia orghelda or potaciones are gener-
ally found in die towns, but their rural echoes are undeniable. Here are the
beginnings of those rural solidarities which were to be one of the features of
rural life so visible in die twelfth and thirteenth centuries. Perhaps because of a
shortage of sources elsewhere, we do not find rural communities outside Spain
before about 1020; their early development in Spain was no doubt favoured by
the incipient reconquista. Was it, at the beginning, merely a question of groups of
fideles fighting against Islam ( sagreres ), or was it already a matter of agreements
amongst peasants ( consejos )? In any case, the first allusions to franque^as can be
found from 975-80 onwards in Catalonia, around Cerdana. There one finds
jurados (perhaps elected?), who supervise good justice, and paciarii, who main-
tain public order and watch over die observance of custom. These are not yet
fueros, die forsoi the period after 1050, but the use of the term burgenses around
1020 to designate these peasants - armed, it should be noted - says much about
the stages which had already been passed through. When die local church has
become a place of refuge and the houses, now regrouped, are surrounded by a
palisade (Etter in the German-speaking regions), the group of free peasants
can emerge from the shades of anonymity: a text from die Maconnais of 928
calls them the meliorparsd

5 Recueil des chartes de I’abbaye de Clunj, ed. A. Bernard and A. Bruel, n, no. 1 240, p. 328.

Cambridge Histories Online © Cambridge University Press, 2008

Rural economy and country life

THE PLACE OF THE DEPENDANT

Responding to a question posed by a count, an officer of die Carolingian court
declared drat ‘there are only two kinds of men, free and slave (serous )’. 6 But in
1042 a text from the archives of Cluny could speak of ‘two free men, of whom
one is a serf (j terms )’ 7 At the very core of the subject being treated here, there is
obscurity and confusion. Historians cannot rescue diemselves by talking of
certain men as ‘half-free’: liberty is indivisible, though one may possess it in a
different sense from one’s neighbour. To make it still more difficult, it is pre-
cisely between 900 and 1020 that an essential characteristic of unfreedom
emerged. Let us try to see more clearly.

First of all, it should be said that there is no doubt diat real personal servi-
tude, that is to say animals with human faces, continued to exist. This wonder-
ful heritage of antiquity, which Byzantium and Islam were to revive cheerfully,
remained solid, even if Africans and Goths had given way to Slavs (who from
Charlemagne’s time gave the institution their name) or Scandinavians. The
church condemned die institution, indeed, but very gently: many of its digni-
taries, especially in southern Europe, fed and exploited these human cattle.
Anyway, she did not admit slaves to her own ranks, and preached acceptance of
one’s lot in this world in view of the world to come, while at the same time
denouncing the Jews guilty of sustaining the trade. It is difficult to estimate the
size of this rural labour force: serving-women and concubines, carters, or
seamstresses in the women’s workshops, were carefully and separately spec-
ified, and some polyptichs, moreover, distinguish mancipia (a neuter noun) from
other subject persons, servi. Their presence is visible only in the occasional allu-
sion to the slave trade, at Cambrai, Verdun, Magdeburg or Chur, where in the
tenth century die bishop still levied a tax on the sale of any person. Beyond
the Channel, those who were shipped at Hull and Bristol around the end of the
tenth century, for destinations in Scandinavia and the Islamic world, were
probably Welsh and Irish people, boys and girls ‘fattened up’ before sale.
Further south, in Lombardy and the Iberian peninsula around Venice and
Barcelona, the embarcation-points for die Islamic world, slaves undoubtedly
formed the bulk of the peasants and shepherds working on the great estates.
These were real slave-gangs, and rebelled as such, for example in Leon in 975
and in Lombardy in 980, before they were savagely put down as in all slave
wars. True, there were manumissions, either by charter, following the ritual of
antiquity (per cartulam), or by the more recent ritual of die penny placed on die
head (per denarium). There are examples of this in Catalonia around 950, in
Provence around 960, in Leon in 985 and in Lombardy around 1000, but it

6 MGH Cap I, no. 5 8, p. 145 : Responsa misso cuidam data , c. 1 .

7 Recueil des chartes de l’ abbaye de Cluny, ed. A. Bernard and A. Bruel, iv, no. 3380, pp. 475—7.

Cambridge Histories Online © Cambridge University Press, 2008

ROBERT FOSSIER

should be recognised that it was really the Christianisation of the Slavs and
Hungarians before or around the millennium, coupled with the economic
difficulties of the Byzantine empire after the death of Basil II and of the
Islamic world faced by Seljuk attacks, which were the real causes of the decline
of slavery. Moreover, dais was never final: slavery continued in existence.

These beings deprived of all rights were not the only bondmen, as they were
known in Anglo-Saxon England. Numerous others are found in die texts,
people whom a long historiographical tradition has agreed to call ‘serfs’, even
though the word servus is not necessarily the commonest term used by con-
temporary sources, which prefer more complicated but more precise designa-
tions such as homines de corpore, homines de capite, homines proprii or homines cotidiani
(terms all making clear their dependence on a lord), or others stressing their
subordinate role, such as manuales , bordarii or Hausdiener. Their existence and
status were the subject of excited debates among the historians of the 1950s.
Nowadays we are less concerned with the issue, primarily because they did not
make up a very large part of the population. In Bavaria their share has been
estimated at 18% around 1030, and this would appear to be on the high side;
there were very few in Italy and Spain, or further north in Normandy or
Picardy, though they were more numerous in central France and beyond the
Channel and the Rhine. It seems clear that their appearance, from die end of
the eighth century at the latest, was the product of a number of phenomena,
though these themselves are not easy to study: former domestic slaves housed
( casati ) on a holding who gradually freed themselves; freemen who had volun-
tarily entered servitude to secure protection; domestic personnel ( stipendarii ,
nutriti ) whose humiliating position had caused them to be regarded as unfree;
tenants who had become incapable of meeting their obligations and so
excluded from freedom. It seems certain that the transformations of lordship
of the tenth and early eleventh centuries accelerated the process. These
people were, first of all, economically affected; they would henceforth be
expected to do days of labour service, and the payments by the next genera-
tion for succession to their holdings would allow lords to reclaim the whole or
a part of the wealth they had accumulated. Even their marriages, because of
the consequences for their succession, could be supervised and taxed ( marita -
gium , merchet), although these constraints do not seem to have been either
established or typical. In the end their lot - their reduced quantity of freedom,
the ‘stain’, the pensum servitutis which would be invoked from around 1000 in
Italy - was undoubtedly characterised by exclusion and mistrust: they did not
bear arms or attend public courts, were pursued with hounds if they fled and
chastised if they committed a crime, were rejected by women of a different
status, were confined to one corner in die church. They were not beasts - they
were baptised, could possess moveable property and have skills - but it is clear

Cambridge Histories Online © Cambridge University Press, 2008

Rural economy and country life 39

that ‘servitude’ was a pillar of die audiority of lords. To diese we must now
turn.

THE RICH

It is an established tradition of European historiography to concentrate on die
small handful of the very powerful. There are two powerful explanations for
this persistent distortion of our view of the past. First, almost die whole of our
written documentation, and often a large part of the archaeological evidence,
tells us about them. Medieval history has for a long time appeared as a tedious
sequence of trivial conflicts between lords and clerics. The second, more sig-
nificant reason is that this tenth of the population ruled over the others and
determined their destiny, especially in die countryside. Their problems have
come to flood the history of the period: ‘feudal society’ or ‘feudalism’ are the
terms used. Let us try to sort out die essentials, beyond what has already been
said when we discussed the family.

Wealth at this time undoubtedly meant land; those who owned large chunks
of it ruled over odiers. It is practically impossible to make any assessment of
the size of these great estates before about 1050-80; even those of the church
evade any estimate. True, the hundreds, of thousands of hectares possessed by
the great monasteries of the early middle ages had been partially dispersed, but
it has been suggested that ecclesiastical lands amounted to about 25-30% of
the total, and that public property and the lands of the warrior aristocracy
amounted togedier to about as much again. The slaves and tenants just dis-
cussed lived on these estates, which were generally exploited indirectly: diese
were to form the basis of the seigneuries still being established. But at this time
it was the ties between men which were of greatest significance, and which
wove lineage solidarities on the one hand and the great mass of dependants
and servants on the other into die familia , the word used to refer to the collec-
tivity of diose who lived around and were dependent on a lord.

The formation of a loyal but greedy clientele around the rich, who expected
aid and counsel from it, goes a long way back and is an inherent characteristic
of an inegalitarian society. At the time we are considering, the difficulties of
subsistence and the dangers of the environment could only lead to a general
spread of such accretions of amici, parentes and homines around anyone owning
significant granaries. If besides this he was also invested with some public
function, even if diis was only theoretical, the pressures making for such an
accretion would only be the greater. The presence in or near to the lord’s resi-
dence of nutriti or prebendarii, dependants or impoverished relatives whom the
lord sustained, familiares , criados , gasindi, geneats, to cite a few varied terms,
charged with guarding the lord or some odier task, created a familial

Cambridge Histories Online © Cambridge University Press, 2008

ROBERT FOSSIER

atmosphere which has led to the description of this whole aristocratic ambi-
ence as ‘smelling of tire household loaf’. This complex served as a basis for at
least three elements.

The man who fed others and could protect them (the ‘giver of bread’, the
Saxon hlaford), whose riches translated into generous presents and favours and
an open table, lived nobiliter, that is to say without calculating, giving openly,
even wastefully. He who did not was thus ignobilis, ignoble, as was the case with
the pauper or the merchant. The immediate problem is thus the nature of
‘nobility’, which has so greatly divided historians. Some see in it a supreme
group, the only one to enjoy all the elements of liberty, even in the face of
public authority where diis existed. Others have supposed that in the tenth
century it was the blood link with the Carolingians which alone conferred
nobility; some have established a link with a real or supposed devolution of
public authority. But it is generally agreed that, in this period at least, nobility
was an indication of pure blood which was kept in being by a systematic
endogamy practised in spite of die efforts of die church to break up its rivals.
That did not make every great landowner automatically a ‘noble’, but he could
live like one, could aspire to become one, and nothing stops us assuming that
his peasants knew something about this too.

By contrast, the establishing of firm personal ties between lord and depen-
dants did not have to take account of this criterion of ‘nobility’. One is sur-
rounded by those commended to one because it is better to keep them on a
short rein. We know, moreover, that the Carolingians actively encouraged these
practices, which were old established but seemed to diem a means of moulding
society more closely around them. The rites of vassalage are known from the
end of the eighth century, and throughout our period they survived and spread.
It should be noted, however, that they were still not clearly fixed, for in 1020
Bishop Fulbert of Chartres was to explain to the duke of Aquitaine the duties -
all negative, incidentally — which he could expect from homage. 8 Naturally, it is
the material counterpart to this engagement which is of concern to us, because
the commended person, having become the man and hence in theory the equal
of die more powerful lord, had to perform tasks (diese were still called opera,
‘works’, in Saxony in 93 (?) to justify the gift received. At any rate this was a fre-
quent arrangement, though vassalage without a material counterpart is still
clearly visible in Germany around 1020, and equally we find grants of land
widiout homage in Italy around the same time. Such grants were also old-estab-
lished usage, simple temporary loans of land ( laen , Lehn,prestimonio), then per-
manent concessions soon to become hereditary. This is not the place to survey
die development of ‘feudalism’ and the distortions which marked it from 1020
onwards, but its role in reinforcing die aristocratic group at a time when, as we

8 Fulbert, ep. 51. 9 Ganshof (195 5), p. 71.

Cambridge Histories Online © Cambridge University Press, 2008

Rural economy and country life

have seen, the peasantry were beginning to form their own solidarities made a
significant contribution to the hardening of rural society.

This was especially true if die bearing of arms was to become the virtual
monopoly of a restricted group. The idea that every free man was a soldier had
never vanished. Beyond the Channel, the Anglo-Saxon Jyrd was still not seri-
ously shaken; but on die continent more and more use was made of heavy
cavalry, which excluded the peasant and reduced him to the level of a subsidi-
ary force, patrol, watch or substitute. Henceforth the soldier par excellence , the
miles, would be the man on horseback, the chevalier os Ritter. But the Germanic
languages preserved die domestic origins of such men: Knecht (i.e. servant),
knight. The familia of the rich contained enough vigorous boys to make good
knights. These were the people armed to defend die lord, though at first it was
not necessary to make them into vassals or choose only the noble for the
purpose. The milites, who appear from about 920—50 in soudiern Europe and
from about 980-1000 in die north, were soldiers in the making, fed, equipped
and lodged in their lord’s residence. In Germany diey were even recruited from
among the ranks of the servile. Because of the need for convenient access to
the services expected of them and the cost of their arms it was self-evident
that they were casati, garrisoned, and that diey had to do homage. This develop-
ment, which came to mean diat the prestige of the warrior, that of one who
had joined an elite militia after die magical ceremony of dubbing, was so great
that a noble would no longer refuse it and would even strive after it, is already
visible, but these elements were not to fuse until around 1100, and in some
places even later; around the year 1000 they were still unquestionably distinct.

The study of rural society, which is to say of almost die whole of society, has
of necessity taken us to the edge of scholarly fields which need further discus-
sion. A general survey of human society was needed. It will have been noted,
finally, that if die inequalities of wealth, rights and power were very strongly
marked, the general environment widiin which all social levels operated had a
certain homogeneity. The main reason which can be given for this is that every-
one in our period was engaged in what I have called a process of regrouping
( encellulemeni ), a process which seems to me the most important feature of the
break marked in European history by the millennium.

ENCELL U LEM ENT 1 0

There is a solid European historiographical tradition which sees in the coun-
tryside of fields and villages which still surrounds us an ancient and even a

10 Translator’s note: the French term has been retained here and elsewhere to describe the social, eco-
nomic and political process of human regrouping accompanied by a restriction of horizons which
divided society up into ‘cells’.

Cambridge Histories Online © Cambridge University Press, 2008

ROBERT FOSSIER

natural state of affairs. Current upheavals within rural society should, however,
make us reflect on whether similar transformations may not have taken place in
more distant periods. To put it another way: the ‘unchanging serenity of the
fields’ and ‘the eternal village’ are figures of the mind. Historians of antiquity,
seduced by the disappearance of towns or of some villae, have hardly asked
themselves about the state of the Roman countryside, and this is even more
true of the lands beyond the Rhine-Danube limes. Supported by archaeology, it
can be said that an organised field system, that is to say one underpinned by a
network of paths, and the village itself, are creations of tire European middle
ages, and that it is precisely in the period before tire millennium that we can find
tire first signs of it.

NEW FORMS OF LAND-MANAGEMENT

Since even excavations do not provide an indisputable image of what was
going on before tire millennium, textual evidence ought to alert the historian to
tire changes taking place amongst human groupings. The first sign of this is the
appearance of new terms to denote equally new forms of exploitation of the
soil: cella and curtis decline, and even villa tends to take on more of the meaning
of ‘village’; mansus persists, as does hide, but the words lose their association
with obligations and come to mean merely ‘a holding with a house’. Terms
denoting fragments of land - sors, massa, quarterium, area, locus — follow close
behind, and these features are evidently very different from the mansionak ,
villare and casale of an earlier period. At the same time unambiguous expres-
sions underline the movement of population: congregatio hominum, instauratio ten-
imentorum. In short, there are obvious signs of a transition from the former
fluidity in the rural habitat to the framework shaping the rural life with which
we are familiar today. These are signs of a regrouping and a taking control of
men, reassembled into fixed points within the cells of the seigneurie, a process
which I have termed encellulement.

Besides the general causes for the upheavals which Europe knew at this
time, spoken of earlier, various explanations have been offered for these con-
centrations of population. The decline of tribal wars, for example, and the
turning by warriors to more local horizons have been used to explain the quest
for authority and profit which required a closer control over the inhabitants of
tire countryside. So has a decline in rudimentary agricultural methods such as
gathering, shifting animal husbandry and long-fallow cultivation, which imply
a fixing of the cultivated area and a more determined exploitation of the uncul-
tivated area. It has also been suggested that the evolution of family structure
was accompanied by the presence or the persistence of human groupings such
as the hundred ( centena , hundredd) or even simply parishes. All these possible

Cambridge Histories Online © Cambridge University Press, 2008

Rural economy and country life

explanations - even though the prime mover in the chain of causation would
still need to be identified - are not mutually exclusive, and none of them will be
privileged here. But it will be necessary to examine the clearest signs of a move-
ment whose beginnings are perhaps found between 920 and 950 on the south-
ern flank of Europe, between 980 and 1010 in the region from the Atlantic
coast to the Rhine, and still later beyond the Channel and die Rhine, even if dais
periodisation is perhaps principally determined by the survival of sources.

If the general tendency towards a disaggregation of die great domains of
the early middle ages is borne in mind alongside the effects we have just seen, it
seems certain that, by contrast with these elements of disintegration, there was
a hard core of demesne lands which resisted all tendencies towards dismem-
berment, and there were even powerful trends towards the accumulation of
lands, especially in the hands of churches and die chief holders of banal
powers. Ecclesiastical documents, for example, show the abbey of St
Emmeram in Regensburg holding 21% of its lands in demesne between 1000
and 1030; in England, the figures for the abbey of Burton and the bishopric of
Winchester are 40% and 22% respectively. Establishments like Farfa or Monte
Cassino in Italy, Seo d’Urgel or Liebana in Spain and Saint-Amand or Saint-
Bavo in Flanders largely succeeded in reconstructing their patrimonies, often
at the expense of allodial peasants who sought protection from these monas-
teries. As far as we can trace their activities, lay magnates did the same: in
Catalonia, Provence and Latium, where documents reveal their activities after
950, there were substantial concentrations of estates ( congregatio fundoruni).
Economic motives evidently lie at the root of this, for die appetite of die rich
was directed towards soils with good yields or tithes providing a reliable
income which was stopped by uncanonical means from reaching its intended
recipient. The church, of course, legislated against this (the councils of Trosly
and Coblenz, in 909 and 922 respectively) or protested (Ingelheim and Saint-
Denis, in 948 and 992 respectively) or threatened (Seligenstadt in 1022); but in
vain.

The other side of this process, the disintegration of large estates, can be
fairly precisely dated by die development of acts recording sales or exchanges
of lands between laymen and the church, which is evidendy a more striking
sign of a search for profit than the decrease of gifts made in alms; the effect of
the latter was no doubt much the same, but the spiritual component can distort
our judgement. In almost all the regions where it has been possible to count
such things, the peak of change appears to fall between 950 and 1025. This is
true of the changes in alotissements n in Lotharingia, of gifts in Germany, of the
dissolving of contracts of aprisio in Languedoc. It is difficult, especially given

11 Allocations of small parcels of land.

Cambridge Histories Online © Cambridge University Press, 2008

ROBERT FOSSIER

the details, which get in the way, to follow the broad trend of exploitation in
different regions each with its own peculiarities. We may simplify by distin-
guishing between three large zones with different trends.

England, the Seine basin and its neighbouring regions, the main part of
Lotharingia and Germany displayed two related trends. The first is die weaken-
ing of the ties, which in these regions had been strong, between dependent ten-
ancies and the remaining lands held in demesne. The Villikationsverfassung, to
use die German term (die ‘manorial system’ of English historians), began to
break down, especially on its edges, where more distant centres gained their
autonomy. One of the human consequences of this relaxation was to cause the
lord’s hand to fall more heavily on those peasants remaining under his control.
From Dijon to Lorsch, from Saint-Bertin to Regensburg, the tenants close at
hand were severely exploited, while their counterparts furdier away largely
freed themselves. The other feature is the division of die unit of exploitation
into two (. Halbhufe ) or four (' Viertel quartier verged), or even eight, as in England
(bovate). The typical holding shrank from io-i 2 hectares to 3 or 4 , and the new
terms which appear, croada in Lotharingia (from corvee ?) or the boel imported by
the Scandinavians, seem to imply the same size.

Southern France and northern Spain, where the links between the different
parts of die estate had always been loose, followed a different route. The initial
core, the mas doumenc, the domenicatura, lost control over oudying holdings. Since
the dependent holdings in this zone seem to have had single tenants and not to
have been distributed in parcels across the fields as they were further nordi,
each of the mas thus liberated was able to form a new litde unit. There was
often a survival of a render ( tasca , agriere), which recalled die ancient domanial
link, and this can be seen in words like condamina\\ T h\ch reek of dependency, but
these are mere fossils. Besides, the comparatively dispersed nature of setde-
ment and the extent of uncultivated lands in those zones not much favoured by
nature allowed the expansion of these isolated mas, often by usurpation, up to a
size of several dozen hectares.

Italy remained a special case, even if we disregard the contrast between the
Lombard plain and the rest of the peninsula. Here the curds held out against
disintegration in some areas, but two elements shook this coherence: die leases
granted perlibello to the peasants gave them a lot of elbow-room, largely to their
benefit, if their holding ( sors ) was not in the immediate ambit of the cartes', and
in Latium if not in Lombardy die phenomenon of encellulement (here known as
incastellamento ), which here took precocious and powerful forms, broke up the
domanial network more completely than in any odier region.

These varied developments had important consequences for the general
condition of dependants. The loosening of ties with the demesne affected ser-
vices first of all, especially day-works and plough-works. The time was near

Cambridge Histories Online © Cambridge University Press, 2008

Rural economy and country life

when the lord, tired of seeing these performed badly or not at all, would have
them commuted for a money payment, liberating the well-off peasant and
crushing the poorer one. Then the rents which custom would gradually fix at
an unchangeable level became divided into two parts: a render in kind or in
silver (once it had begun to circulate again) at a bearable level, or else a portion
of the harvest ( tasca , champart) whose interest for the tenant lay in the possibility
of escaping from the consequence of climatic fluctuations, so that he would
try to make this a more general practice from about 1020-30, especially on
newly cultivated lands. In this context we must also note that the subdivision of
peasant tenures reached a new low level, around 3 or 4 hectares (though still
with immense variations - in Catalonia around 1050 between 1 and 1 9 hectares,
for example!). This situation can be explained in two ways: either, and this is the
optimistic view, technical progress meant that 10 hectares were no longer
required to feed a family, or else demographic growth and the evolution of the
family proceeded so fast that they forced the break-up and an overloading of
tenements.

We can now see why we needed to make this survey of cultivated lands
before examining tire environment. The allodialists, whether large or small,
who continued to direct their exploitation of the land, and the tenants, over-
crowded or not, who were freed from ‘demesne’ constraints, formed a mobile
mass, juridically freer and available to be regrouped. True, powerful owners
already possessing their ‘men’ or even their slaves continued to exist. Equally,
the disintegration of the ‘system’ had its negative aspect, for example the wors-
ening of the lot of the poorest. But the general effect of encellulement was posi-
tive.

THE REORGANISATION OF THE ENVIRONMENT.

The floating mass of the peasantry had hitherto lacked centres around which
they could crystallise the disparate and disorganised environment inherited
from the preceding centuries. It was not enough to scatter their huts haphaz-
ardly within a clearing, not even enough to give a name to this agglomeration;
there had to be a coherent organisation there, so that a state of mind could take
hold. Augustine said that what made a town was ‘not walls, but minds’; some-
tiling similar could be said of the countryside.

It seems to me to start with that it was the dead who fixed the living. The
ancient necropolises were laid out along the roads leaving towns, and those of
the early medieval countryside on die edges of the lands between settlements,
always far from housing, perhaps because of a fear of the dead which the
Christian church would slowly uproot, though hardly before 900. Thencefordi,
especially as regards magnates, it was, as has been said, psychologically very

Cambridge Histories Online © Cambridge University Press, 2008

ROBERT FOSSIER

important to live alongside one’s ancestors, to respect them, even to consult
them. And since it was hardly easy to move them about, diere was a tendency to
group setdement around tombs; archaeology has shown, for die whole of
northern and north-western Europe at least, that the cemeteries and the pro-
tected areas surrounding diem which enjoyed peace and functioned as assem-
bly-grounds [atria), and are still in use today, date from the tenth century, not
before. Indeed, the necropolises of the preceding centuries, set in the open
countryside, were abandoned at this time: the cemeteries of Normandy or
Wiirttemberg were no longer used after 850 or 900. Unfortunately for histo-
rians, the accompanying disappearance of funerary goods and the practice of
burying in shrouds of perishable material often mean that dating these new
village cemeteries is difficult. Where it has been possible it is by no means
uncommon to find diat the field of the dead, the ‘second village’, is earlier than
the parish church, as in Levezou or in lower Saxony. Moreover, die council of
Tribur (895) enjoined the separation of church and cemetery, and if that of
Toul (971) prescribed the establishing of a cemetery in the middle of each
Christian village, it did not require it to be located next to the church.

It is evidently the latter which came to be the heart of the new village, so
much so that right across Europe it still symbolises the rural settlement. It is
not our purpose here to discuss the slowness with which a parochial system
was established, once die Mediterranean shores had been left behind. It is suffi-
cient to recall that around 920 die diocese of Paderborn, which amounted to
about 3000 km 2 , had no more than twenty-nine parishes, and that there were
no more than 3500 in the whole of Germany. As for church buildings diem-
selves, even if there is no shortage of buildings whose foundations are older
than die tenth century, nothing shows that they acted as a focal point for settle-
ment in that earlier period. The example of the fana of antiquity or of Christian
oracula in die open countryside are enough to show that this was not necessary.
On die other hand, we know that rural Christianisation often took the form -
in Gaul or Saxony, for example - of establishing a baptismal font at the centre
of a fundus, an isolated great estate; the fact that this practice allowed the devel-
opment of the proprietary church ( Eigenkirche ), with implications evidently
counter to the spirit of canon law, is not our problem here. What seems most
clearly established is the breaking-up of the giant rural parishes of previous
centuries (die plebes cum oraculis ) into more modest units capable of stabilising a
small group of die faithful. This phenomenon has been noted for the pievi in
Latium between 9 5 o and 1 020, in Auvergne and Poitou before 1050, where the
word parrochia comes to replace the word villa as the term for a nucleated settle-
ment, and later on further north. By and large the former kernel preserved a
certain primacy over the subsidiary units established within its initial territory,
and the present-day parish map often still shows this; but sometimes the

Cambridge Histories Online © Cambridge University Press, 2008

Rural economy and country life

churches established in the villages on the uplands led to the remorseless aban-
doning lower down of the old parochial core church (Niederkirchi).

If the bell-tower around which the members of the parish came together
was a symbol of the village, the casde itself was the symbol of the middle ages.
But if for now I merely touch on it, it is because I drink it developed later than
the other two crystallisation points for the reshaping of setdement, where it
developed at all - something which justifies our restricting the concept of incas-
tellamento to those regions where its role is clearly established. To set up a for-
tified location for assemblies protecting and exploiting men is a phenomenon
found in all ages. For the centuries immediately preceding those we are here
concerned with, archaeology has clearly revealed both the ancient and revived
oppida which still served as royal palaces in Germany in the tenth century (as at
Werla and Tilleda, but also further west), and also the huge earthworks of a
more recent age (some seventh- or eighth-, some tenth-century) found in the
Auvergne, in Normandy, in England or in the Palatinate ( Ringwallen , ringworks,
etc.). The original feature of the tenth century was the way in which Europe
came to bristle with strengthened buildings, towers first of wood and then
from the end of the tenth century of stone [turns, dunio), set up on a natural or
man-made elevated site, surrounded by a moat and possibly a protective enclo-
sure (bailey), and designated by revealing terms - motte , rocca, podia, colli —
echoed in modern terms like Wasserburg and ‘moated site’. Their location was
rarely chosen at random: they were set up at an ancient assembly-ground
(Maine, Oxfordshire), a Roman mansio (Piedmont, Burgundy), a villa or casa
(Lombardy, Auvergne, Rhineland), or a cult site (in the Liegeois). Such loca-
tions, especially when we note that there was a strategically superior site nearby
which could have been chosen instead, show that the aim was much more one
of economic surveillance and of social control than of military utilisation.

The material and judicial status of these constructions is fairly well known.
Everywhere where public authority retained its force such towers were built
only with permission, generally from the count, with or without a genuine dev-
olution of powers in the form of regalia. Usurpations by daring allodialists were
not unknown, but they rarely survived without either punishment or, more fre-
quently, retrospective legitimisation. Anyway, it could also come about that the
rights of control over the men of a neighbouring village [mandamentum, salva-
mentum, potestas) were granted subsequently, either to a landowner who had
behaved himself, here functioning as a deputy, or to a military leader whose
support was needed [castlania in Languedoc and Catalonia). The very force of
things meant that possession of a tower, especially where no effective public
power could watch over the protection of the villagers, implied tire possibility
of gaining jurisdiction over them ( districtus ) or exploiting them ( feorum in
England); this was the core of the seigneurial unit. One can thus understand

Cambridge Histories Online © Cambridge University Press, 2008

ROBERT FOSSIER

the ties between the powerful man and those without whose aid he could not
have built his tower: it has been calculated that forty labourers would have
needed fifty days’ work to put up a small round motte io metres in height and
30 metres across.

The rhythms of growth, which can be revealed only by archaeology (textual
evidence being largely absent), are also fairly well known. Once again it was
southern Europe which set the pace, perhaps because of the disappearance of
higher authority and because of local disturbances and conquests. In Italy the
movement began in the peninsula around 920, and around 960-70 in the
Lombard plain, with about 120 casdes in Sabina by 1050, for example; in
Catalonia, with the help of die reconquista, the starting point was 950-60, with
nearly seventy towers by 1025 ; in Provence and Languedoc it was between 980
and 1020, often on fiscal lands, with a hundred castles between Luberon and
Costieres before 1030; by diis last date some 1 50 castles had also been built in
the Massif Central since 970—80. The further north one goes, the later the
beginning: in Poitou it was around 980, but with only fifteen castles being built
before 1020, as in die Maconnais. North of the Loire the movement did not get
going until after the millennium, in Anjou and Normandy hardly before
1030-40. The wave crossed the channel with the Conqueror from about 1070,
and beyond the Seine everything changes around 1060-80. In the Low
Countries, Lotharingia and the regions beyond die Rhine it was closer to 1100,
well beyond the temporal limits of this chapter.

Of course there was a gap between the building of the casde and develop-
ment of an accompanying control over men. Sometimes its lord was able to
use force to regroup the peasantry around his rocca in a castro or castelnau, a
development favoured by geography in Italy, Provence, Gascony or Catalonia.
Surrounded by professional horsemen ( caballarii castn ) he was able to control
the fortification, carry out police duties and summon before his court at least
those cases involving lesser justice. As castlan or castellanus he was the protector
and the lord of custom ( consuetudo castri, ius munitionis). The relative solidarity of
the peasantry in these parts may have forced him to behave more circum-
spectly: in Auvergne and the Languedoc he will have used persuasion, promis-
ing benefits to those peasants who came to populate die barns which linked
together at the foot of his walls. A better way was to attract artisans whose
work would maintain die equipment of the group living in the casde and who
soon, headed by the smith (faber , fevre.ferrario), would give a lead to the rustics.
Further north violence was less customary, since dukes, counts, kings and
emperors were not just vague memories in these regions. Frequently the
regrouping of men preceded the appearance of the castle, which would arise a
generation later all the richer and more powerful. In Burgundy, villages were
formed by the spontaneous grouping of inhabited mansi (here called meix ), the

Cambridge Histories Online © Cambridge University Press, 2008

Rural economy and country life

kernel of a community which would soon gain self-consciousness; in Picardy
and Wurttemberg the palisade ( angle , Ettet) which surrounded the setdement
shows that it came first, and here the casde did not engulf the village, but rather
looked out over it. In England, die association between dwellings and markets,
the distance of die casde and powerful constraints of public or communal
obligation (such as armed service in the jyrd in Anglo-Saxon England) charac-
terised this slow development. In the end, however, it undermined overall
freedom.

BOUNDARIES AND HOUSES

It would have been good to conclude this attempt to classify setdements by
saying something on the problems posed by their external appearance. But
here we have far more questions than answers. It is much the same with the
state of die cultivated area over which the villages, once formed, extended their
control and exploitation. To estimate the extent to which they mastered it we
would have to be able to say that the full network of roads and paths was in
place. Here archaeology is powerless: there have been attempts in England,
Alsace or around Limbourg to date either the hedgerows or the fields
{Ackerberg ), but the results are too uncertain for such distant periods. There
remain the texts from the Sabina, from Burgundy and from Catalonia in die
tenth century which mention boundaries. Alas, three times out of four the
scribe mentions the name of neighbours or of a natural feature, and only one
in four a road. The conclusion must be that around the millennium the field
pattern was not yet established, but still in the course of formation. There is
only one exception, of which historians of southern Europe have made a great
deal: the traces of Roman centuriation. A number of authors have claimed that
field layouts were based on the squares of the centuriation system, and have
offered examples from Languedoc and Lombardy. Filled with the desire to
emulate this, others have wanted to see centuriation everywhere, even in those
places where it would be quite irrational to suppose it. Quite apart from the fact
that the architectural remains of antiquity do not coincide with modern cadast-
ers, which are likely to reflect the arrangements of the high middle ages much
more closely, I cannot see anything in these possible coincidences other than
irritating archaisms which testify merely to a tendency to make some use of
what was already at hand.

I cannot any longer avoid a problem often invoked by the partisans of conti-
nuity who wish to play down the significance of tenth-century transforma-
tions: the antiquity of very many of our village names. It is indeed true that die
stock of place-names whose origins are indisputably Celtic, Iberian, Germanic
or Roman is impressive and this could suggest that the habitations they

Cambridge Histories Online © Cambridge University Press, 2008

5°

ROBERT FOSSIER

designate are just as ancient. But I do not believe it: apart from the possibility
that they may merely have designated an isolated element which then served as
the centre of a human concentration once all others had disappeared, I would
think that such terms often designated human groups (as it certainly did with
Roman names ending in -iacum or Germanic names ending in -ing) and moved
around with them, becoming fixed when they did. The baptisms, rebaptisms
and displacements of villages which we can still see today in Europe, and still
more elsewhere, ought to be enough to convince the advocates of perma-
nence.

There is a third problem, which it is particularly unfortunate that we cannot
solve: what did villages and especially houses of the millennium look like?
Unfortunately, the conclusion drawn from die foregoing must be that the vil-
lages and houses of our period lie underneath diose of our own. Although we
have many examples of previous habitations, abandoned in the seventh, eighth
or ninth centuries, we cannot use them to help us. To describe Chalton
(Hampshire), Kootwijk (Guelders), Maizy (Champagne) or Warendorf
(Westfalia) and so many others which had been abandoned before 900 would
have no interest. Those centres whose displacement has been light enough to
allow us to say something — Hohenrode in the Harz or Wharram Percy in
Yorkshire for example - are very rare, and what we have been able to discern
from them is modest. Houses were still large, 8 metres by 10 or 12, with
beamed roofs, perhaps widi die addition of a solar in the case of lordly houses.
These are traces of a family group which was still large; the houses had exterior
doorways and underground foodstores. All the same, change was barely begin-
ning; the task of describing twelfth-century villages can be left to odiers, but
they are evidendy the continuation of developments whose origins may be
traced in the shadows of the ‘dark age’.

THE BIRTH OF THE SEIGNEURIE

We have now arrived at the key result of encellukment. From the tenth to the
eighteendi century it was die seigneurie in which the men of Europe lived, in
forms showing wide chronological and geographical variation. The regions
such as Ireland, Scodand, Frisia, the Basque territories and a few valleys in the
Alps and the Apennines which did not know it were rare. The fact that one can
also show that many towns in western Europe were seigneuries, which is not
our concern here, only serves to underline the importance of the problem.

Having said diis, it is distincdy difficult to trace the means by which the seig-
neurie developed. We certainly cannot understand it widiout taking into con-
sideration all the phenomena we have just discussed, including castles; but a
‘political’ perspective is also required, even if it has its limitations. We know

Cambridge Histories Online © Cambridge University Press, 2008

Rural economy and country life 5 1

that the early middle ages were characterised by a public grouping of men into
territorial units generally known as pagi, whose origins have been much dis-
cussed. At their head stood an official representing the prince - comes , ealdorman,
gastald, Gaugraf, and so on. Around 920—30 one could have listed about forty of
these units in the British Isles, 1 60 in western France (two thirds of them north
of the Loire), twenty in Christian Spain, eighty in Lotharingia, and more than
220 in eastern France; in Italy one might estimate them at perhaps 1 50. At mili-
tary camps, at the centres of walled towns or at palatia, justice was done, fiscal
lands surveyed, free men summoned to arms, and taxation raised if anyone
dared to, or at least imposts for war, service and forage ( Heregeld , fodrum, alberga,
hostilicium, etc.). In practice, however, from the ninth century onwards the
count, who in general had none of the characteristics of an administrator,
turned to a deputy to help him out. In England die shire-reeve or sheriff played
an essential role. Nevertheless, dais delegation was not enough in those cases
where the size of the pagus was too great (was it perhaps a function of the size
of the population?) or where there was need to take rapid decisions, for
example in dealing with Vikings, Hungarians or Saracens, or even, given the
slowness of communications, die need to make on-the-spot assessments of
material needs, or minor problems. For this reason there was a need for more
modest territorial units, grouping at most a few dozen centres of population,
known by terms like centena , vicaria, hundred, ager in Germany, western France,
England, Italy, the Alps, etc. Those exercising the rights of a deputy ( vicecomita -
tus) usually had an official delegation of military, judicial and fiscal duties. But
these could be acquired by people on the spot, and we have found a number of
casde-builders amongst them. After about 940 in Italy, 970 in Catalonia, 990 in
Poitou, 1 o 1 5 in Normandy and England, to mention only a few, these men held
the law-courts. That ‘feudal’ or ‘vassalic’ matters would ultimately also be dealt
with there is a different problem, not relevant here. Serious cases, ‘matters of
blood’, pertained, in principle, to a higher court, such as that of die count.

Around the millennium the situation appears to have been this. Holders of
the hannum (whether lawfully or not), that is to say of the right of pursuit, con-
straint and punishment, came to tolerate, and sometimes to encourage, a
dependent clientele of men who were rich and armed, who were their men or
indeed dieir relatives, in building towers and holding courts. Such castellans, in
southern Europe at least, were to become more or less independent, or else
simply seniores or domini, lords. They could not be prevented from dealing with
law-suits affecting land, the most rewarding ones incidentally, before extending
their grasp to odiers, nor from reclaiming for themselves and their men at arms
the rights of gistum and of fiscal assistance. How can we distinguish between
the imposts they exacted from their peasants, allodialists included, which
derived from their ‘public’ rights (from what the Germans call Landrecht) and

Cambridge Histories Online © Cambridge University Press, 2008

ROBERT FOSSIER

those which came to them from their land-holding power? In all cases we find
them as lords of ‘custom’, which ruled the life of all, free to innovate and
indeed to expropriate by inventing make consuetudines or mals usos, evil customs,
which struck so many peasants simply because what is new is inevitably
wicked.

In those regions where royal authority, even if enfeebled, still had some
force (Anglo-Saxon England, northern France, the western part of the
empire), or where comital power remained strong (Catalonia, Normandy,
Flanders, Saxony) this development was reined in and bearable. Elsewhere it
provoked a lively movement of rejection, which itself deserves to be called
revolutionary. The church placed itself at the head of this movement, since it
was even more menaced by such developments, both in its judicial rights of
immunity and in its enormous landed wealth: the ‘peace of God’, so often pre-
sented as seen through the rose-tinted spectacles of piety. The clerics in fact
broke with the class solidarity linking them to die warriors, under die pretext of
coming to die aid of the pauperes, the inermes, deemed to be as dear to God as
diey themselves. I do not here have to describe die stages of the movement,
from the Aquitanian, Burgundian and Languedocian councils of the period
989-1027 with oaths enforcing a truce, moving on to the oaths sworn before
bishops and princes between 1023 and 1048. What we need to remember for
our purposes here is that the church soon came to terms with the great laymen,
especially after the excesses committed by the peasant bands who had rashly
been encouraged to setde dieir accounts, in Normandy, Leon or Berry for
example. Making use of the protection the church enjoyed over its lands - diat
granted by royal advocacy, for example - it advanced the evolution towards the
seigneurial system, whose birth it had so long retarded. Moreover, the network
of feudal relations, familial interests and political responsibilities put a long-
lasting seal on the rapprochement between the first two of the three orders.

Henceforth the seigneurial cells were in working order in villages and
around castles. As we are now already beyond the period of this book I can
confine myself to these few remarks. It does seem necessary, however, to note
a new feature. Whether their authority was of public or private origin, lords
soon mingled these two notions, so readily elided in the middle ages. A number
of obligations due, strictly speaking, only from those peasants who were
tenants of the demesne soon came to be extended to those who had no such
ties; these included corvees, exacted by right of the ban, but put into effect for
the benefit of the landlord. One could even say in this last case that such dues
involved a concept of ‘banalities’ of a particularly ‘illegal’ kind and hence a
source of protests, and this was not a trivial matter: as late as 1000 or 1020 at
Milan and Brescia a week’s work with oxen was being demanded, as well as two
months’ labour by hand; several days a week were required in England {week-

Cambridge Histories Online © Cambridge University Press, 2008

Rural economy and country life

works), somewhat less on the continent. Doubtless this was less than it had
been in the ninth century, but it was now imposed on everyone. So was another
banal demand, perhaps more justifiable: the tax in recognition of protection to
which the texts give the flattering name of ‘request’ {Bede, questa , rogatio) or the
more accurate one of gistum {alberga,gayta, i.e. (forced) hospitality) or, above all,
that termed ‘exaction’ {tolta, tailk, tonsio).

The ban affected allodialists and engulfed free tenants; serfs were unaffected
because they were the property of an individual and did not, for example, pay
the faille. This was the fiction. In the villages which made up a seigneurie, all in
practice were on the same level vis-a-vis the lord. Divisions between them were
not lacking; these derived in particular from the economic problems with
which all were confronted, and to which I shall now turn.

A STILL FRAGILE ECONOMY

In trying to glean what one can know about the European rural economy of
around the millennium, the first line of enquiry should be directed towards
men’s needs, the only criterion by which we can judge whether their efforts
were adequate to satisfy them. This is a difficult subject for the medieval centu-
ries, and especially for those dealt with here. Of course we could assume that
the levels of consumption which we can deduce from Carolingian documents
continued to be valid one or two hundred years later, but can we be certain?
Moreover, there are suspicions of exaggeration and confusions in the numeri-
cal data we have, which at Corbie for example envisage that those doing labour
services would receive 1.95 kg of bread, a litre of wine, 300 g of legumes, and
100 g of cheese and eggs daily, a somewhat unhealthy and unbalanced diet
amounting to about 6000 calories, possibly a ration for a family rather than an
individual. Chroniclers, miserly with numbers, are still vaguer. When Helgaud
goes into ecstasies over the fact that King Robert allowed beggars to pick up
scraps from below the table, or an epic poem tells of a trencherman eating a
peacock in three mouthfuls, this tells us nothing. Archaeology in its turn is of
litde assistance: the finds from the rubbish-pits of Holstein or Hanover in use
between 800 and 1000 show proportions of animal bones consumed which
suggest a high intensity of animal raising, though the data are random and
varied: 1 5-66% oxen, 10-70% swine, 1 1-2 3% sheep. Sparse and late data from
Germany have led to an estimate of about 2200 calories daily provided by
cereals, though the tenth century has also been described as ‘full of beans’.
How, too, should we count the roots from the woods, the eggs from the farm-
yard, the honey from the beehives? In short, none of our data provides any
certain indications. What is more or less sure, as has already been said, is that
there were acute famines in the early eleventh century. The problem can be

Cambridge Histories Online © Cambridge University Press, 2008

ROBERT FOSSIER

summed up in one question, a crucial one: did progress keep pace with the
rhythm of needs?

AN ENLARGED CULTURAL HORIZON

In answering yes to this question I am in effect advancing a conclusion which
only tire succeeding centuries will justify; but such an opening position is indis-
pensable for the observations which follow. Consider in the first place the
subject of tools, which some have wished to make one of the most important
causes of the rise of Europe. Whereas tire middle ages were once credited with
inventions, there has more recendy been a swing to the opposite extreme of
only allowing it a certain talent for popularising others’ inventions. This empty
dispute hardly takes account of a geographic reality: the Graeco-Roman civil-
isation of antiquity had brought to a high degree of perfection techniques for
dealing with wood, stone and textiles, and ignored water, which in its zone was
unreliable, and iron, which was there rare. Central and northern Europe was in
a different position; and besides this, the variety of species diere permitted
progress in the exploitation of animals. To confine myself to what seems
uncontroversial, I shall look at three key areas. Historians first claimed that the
shoulder-collar for the horse or the breast-yoke for the ox had saved Europe.
They then maintained that these techniques had been known to the ancients,
just like the hipposandals designed to protect die feet of die horse. In reality,
however, the iconography of harness and the archaeology of metal parts have
revealed no trace of these practices until the end of the tenth century at the
earliest, and in regions distant from the Mediterranean, such as Trier, Savoy or
Bohemia. Perhaps die novelty of these things, if it was one, came from the
choice of a breed of animal more appropriate to such practices dian those of
southern Europe.

A second point: iron featured greatly in this equipment, as in mills, from a
very early point. Here too we touch on a key sector of medieval technology.
Smithies were particularly numerous and relatively easy of access in the
Pyrenees, the Rhineland and Saxon regions, Normandy and the north of the
British Isles, Burgundy and Champagne. It has also been possible to show that
Germanic or Celtic smithery was well in advance of its Mediterranean counter-
part: more solid axes, ploughshares, coulters and mouldboards, horseshoes,
barrel-hoops and wheel-rims, and of course the armaments used for warfare
and hunting. The man who worked diem in the midst of the sparks and
bellows was indeed the key worker of the village, and its lord was his first and
most admiring client. The exploitation of mines and the practice of smelting in
low furnaces are found at Canigou from 945, around Fulda and Lorsch about
960, in Poitou and Normandy from 975 or 980, in the Ardennes, in the forest of

Cambridge Histories Online © Cambridge University Press, 2008

Rural economy and country life

Othe in die Champagne and in Yorkshire just after 1000. Before 1030 we can
find ‘pigs’ of worked iron entering into lists of tolls, often heavily taxed as at
Pisa or Arras; and increasing numbers of smiths in villages in Catalonia, Sabina
and Picardy between 1010 and 1030 are a sign of die growth, from now on
unstoppable, of a metallurgy systematically ignored by antiquity. A point
should be noted, however, which bears on what was said previously about the
reordering of men: furnaces and anvils were first of all used in the woods, near
to sources of fuel (there are references to coal in Saxony and near Leicester at
the end of the tenth century). In order to make his work more effective,
however, the smidi moved from the forest to the village, and I would readily
claim that the smithy, just as much as die casde, was a crystallising point for the
population.

The appearance of a mill for hammering iron in Germany around 987 leads
us on to the third area of growth: the harnessing of water power. Litde utilised
by die ancients, but technologically very relevant to the needs at hand, the
water-mill became the first European ‘machine’. The regularity of water-flows
above a certain latitude, the fortunate nutritional consequences of the fish-
ponds that were full of fish (the by-product of mill-sluices), die gain in time
and of profit (diough this was later) which the rich were able to draw from the
use of the mill by die people of the village, all explain the staggering success of
these machines. Carolingian texts refer to diem, certainly, but one has the
feeling that their widespread diffusion did not occur much before 920-35: in
Poitou, Catalonia, Berry, the Low Countries. After this period we find them on
every water-course. The effects of dieir installation are known: they were
expensive to build, needing beams and mill-stones of high quality as well as
lead and iron, but they were estimated in die eleventh century to yield revenues
equivalent to those of 20 hectares of land. The rich men who had diem built
thus knew how to get dieir investment repaid by those who used them; those
who could not afford to use diem lost time and energy milling grain at home by
hand. Nevertheless, there is no definite evidence of a ‘banal’ obligation to
grind at the lord’s mill before about 1030-50. It should be added that, if water
was owned and dius had to be paid for, wind was free to all: diere is a reference
to a mill powered by wind in the region of the Spanish Mediterranean coast at
the very end of die tenth century, but it would be a rare phenomenon before
about 1 1 50.

It will have been noted that the major effect of this progress in technology
was to dispossess the artisans of the demesne of dieir former omnipotence.
Weaving, joinery and smithery had been carried out under the direct control of
the lord’s agents, or even, in a kind of ‘wild artisanry’, by those working in the
woods as potters or hermits. Henceforth it would be in the village itself, that is
to say at the foot of the casde, that die transformation of products and

Cambridge Histories Online © Cambridge University Press, 2008

56 ROBERT FOSSIER

materials would be concentrated. This concentration of technology in the
village would allow tire peasantry to apply themselves to the key task of master-
ing uncultivated land, a key to tire extension of culture and the food-supply.

The saltus, outfield, bosc,foresta (probably derived from foris, ‘outside’, not
from Fohre, ‘fir’) was uncultivated, the zone which might be no more than
lightly wooded but which man feared and did not know how to tame. It was the
countryside of the Atlantic seaboard of France, the maquis ( mescla ) or garrigue
of the Mediterranean, the savannah spiked with thorns and isolated trees of
north-western Europe, and of course tire thick woodland of Lotharingia,
Germany and Scandinavia. It rested on poor soils, podsolised and stony, but
also on heavy and potentially fertile clays. To clear it was very hard work; it
extended over regions where animals stronger than man lived, the wolf for
example, or, still worse, those evil spirits who set traps for wanderers. Emperor
Henry IV was lost in it for three days, and in Burgundy the least scarcity could
cause this wild world to spring up again. The analyses of pollen or charcoal
which are today our most reliable indications of the nature and extent of vege-
tation are very striking: in the Ardennes, Hesse, Schleswig, Kent, Bohemia,
Valais, Poitou and Languedoc, for example, we find woodland covering
between 50% and 70% (the latter in Germany) of the surface area, while at the
end of the eleventh century Domesday Book records some 40,000 km 2 of
woodland in England. This huge mass of land was by no means inert or value-
less: its role as a zone for hunting and gathering on its borders and for military
protection or emergency refuge, quite apart from its role as a source of the
principal material from which a wood is made, turned it into a world with a
population of gatherers, charcoal-makers, woodcutters and also brigands, as
well as, above all, of domestic animals left there to pasture freely, even with the
risks which that entailed, because the cleared and cultivated land had to be
completely reserved for growing crops for humans.

But if the needs, of the latter grew, because they were now eating more or
their numbers had grown or their family groups were breaking up, we should
envisage an alternative organisation of the ecosystem. It has even been sup-
posed that in tire beginning it was the needs of animal husbandry that pre-
vented tire peasantry from eroding away the woodland. In any case, the word
‘clearance’, in its primary sense of struggle against the bramble rather than
against the beech, gives a good idea of this struggle, quite modest at first. There
were at least four kinds of attack, no doubt differing in their modalities and
their effects. Heavy soils, marles, limestones and sandstones capable of bearing
good harvests were tackled at the end of the tenth and the beginning of the
eleventh century in tire Auvergne, Burgundy and the Rhineland, the Harz, the
Weald and Sussex. Rather later, waterlogged lands, marshes and coastal zones -
Schorren, moeres, fens - were taken on; these were more suited to the rearing of

Cambridge Histories Online © Cambridge University Press, 2008

Rural economy and country life

sheep. The pebble-strewn floors of the valleys and floodplains of fast
Mediterranean rivers ( varennes,ferragina , rivages, bonifachi) along the coastal plains
of the whole of die southern flank of Europe were probably not attacked until
1000 or 1020. Finally, the exhausting conquest of slopes by terracing may have
appeared from 975 in Catalonia, but would wait until 1025 or later in Provence
or Italy. It goes without saying that where these enterprises were carried
through by individuals more or less illicidy on the lands of others drey largely
escape our observation. We do not find out about them except through the
steps taken by the rich in die form of contracts. The classic example, at least
that which we find in the documentation which the church provides us with,
was the purchase of land by clerics and its putting under cultivation by teams of
lay workers: the quadras of Catalan pioneers, the Barschalken of Bavaria, the sar-
tores of Picardy. The lands thus gained might as a result preserve a particular
status because the ‘guests’ ( hospites ) who had come from more or less distant
places, and been established diere as cultivators with their dwellings, enjoyed
seigneurial protection, personal liberty and fairly light obligations as far as
renders in kind were concerned, as with the gualdi publici of Lombardy, the
lathes of the Weald and the hostalitates of the Pyrenees.

The trend had hardly begun to show its outlines around the millennium, and
it is hopeless to expect to be able to estimate its size at this point. Pollen analy-
sis gives some indications, but no figures. The breaking up of fields into strips
might be a proof of its existence, but when does this date from? As for place-
names, though their evidence is crucial, they cannot normally offer a precise
date. Places ending in -viller, -hof, -dorf, -sart or -bois are perhaps die products of
clearances, but these may go back to tire initial timid Carolingian phase. What
remain are the micro-toponyms which are certainly linked with the struggle
against the saltus. -ley, -den, -hurst and -shot in die British Isles, -rod, -ried and -
schlag'm the Germanic regions, -essart and -rupt in northern France, -artiga in the
Pyrenees, -ronco in Lombardy and many others. But do these date from the
tenth, the eleventh or even the twelfth century? And how are we to decide?

THE STIRRINGS OF PRODUCTION

What we today call the medieval ecosystem, which survived in essence until die
beginning of the twentieth century, was based on a combination of cereal
crops won from the ploughlands, the products of free pasturage in die forest,
and meat and dairy produce, complemented by fruits and roots from the
uncultivated land and the minor products of the farmyard. Clearly, even if
some sort of dietary equilibrium could be attained, this situation constituted,
as has been said, a ‘vicious circle’: to increase the arable at die expense of
woodland was to cut off what the latter could supply, but to preserve the

Cambridge Histories Online © Cambridge University Press, 2008

ROBERT FOSSIER

woodland was to risk underproduction of cereals. This millenary conflict
between the carbohydrates of the plough and the proteins of the forest
depended for a favourable outcome either on improvements in technique or
on an equilibrium of needs. It is unnecessary to say that those alive around the
millennium were not weighed down by the need to find solutions. For many
generations they took advantage of Feldgraswirtschaft, a system of shifting culti-
vation on a more or less regular rhythm of years: clearance, followed by years
of cultivation until the land started to show signs of exhaustion. Such prac-
tices, which die poor could not pursue for lack of sufficient land to do it on,
were the preserve of the lords. This is the origin of the quite untenable belief
in die general existence of die crop rotation which has been seen in
Carolingian documents. In reality we are dealing with an incomplete occupa-
tion of the areas covered by the polyptichs and widi a shifting between winter
or spring grains and a variable fallow period (the famous tres arationes of Saint-
Amand which have so often been cited). Not until die mid-thirteenth century
do we find a conscious and regular rotation; here we are dealing merely with
empirical practices.

What was grown? First of all, cereals for bread-making. The best, grains
yielding white flour, are known everywhere because of the demands of lords.
It has been noted in Catalonia and the Low Countries that after die millennium
the hulled wheats of antiquity such as spelt yielded to a naked wheat which did
not clog up the mills with which the rich were equipping themselves; barley
declined, but beer and oxen saved it; rye resisted because diat is its nature,
robust and plain, quite apart from the quality of its straw; oats, already in use
before 700, source of porridge and soon to be food for horses, begin their
career as a ‘March’ sowing, but are far from equalling the mass of sheaves of
wheat in the granaries. Had panic or millet already made their appearance? The
economic historian would welcome other details as well. And how was work
organised? The wealthy could dispose themselves of the services of their men,
and that in abundance, indeed beyond what was useful: at Brescia there were
60,000 man-days to be used, at Saint-Germain-des-Pres 135,000, which is
surely absurd. But what was expected of them? How much ploughing, what
sowing, what equipment? This last question is crucial, but for the period we are
dealing with unanswerable. We know that the plough of antiquity, the aratrum
with a hardened share, sometimes armed with an iron point, hardly permitted
deep or fruitful ploughing, only light and symmetrical furrows. It remained in
use in southern Europe, but already in the eighth century the Lombards talk of
a ploum (evidently a Latinisation of Pfl/igor plough), which no doubt came from
central Europe, and indeed asymmetrical shares have been found from the
ninth century in Moravia. Unfortunately, the word carruca, which ought to
imply a more effective instrument, appropriate for the attack on rich and heavy

Cambridge Histories Online © Cambridge University Press, 2008

Rural economy and country life

soils, is used by scribes without discrimination. Archaeology has revealed near
Utrecht and in the Belgian Campina fossil fields where it seems that tire two
rival types of plough were in use. What can we say? That the future lay with the
wheeled plough drawn by horses on die best soils? That is certain, once we get
past 1 080-1 100, but before that we can only guess.

Progress obviously has to be judged by results: three elements, none of them
substantial, seem to me to be signs of a beginning. First of all, what is known
about tire layout of fields suggests two tendencies: enclosure, even the provi-
sional kind provided by a brush hedge at the time of sowing, appears to give
way, except in die special cases of vineyards and olive groves, to an open coun-
tryside which can be used for regular pasturage. Besides this, die form of fields
is perhaps beginning to change: even though die massive, almost square shape
(1 quademi and aiole in southern zones) still appears to survive a litde longer, it has
been noted that in the Low Countries and the Rhineland and Bavaria we can
see in outline die beginnings of a system of strip fields, though it is true that
before 1025 this seems to be known only in England (widi parallel solskfts
grouped into quarters or furlongs). This kind of layout, which can only coinci-
dentally be seen as associated with a particular kind of plough, appears to rep-
resent the abandoning of the very primitive technique of crossed furrows used
in antiquity. A second point is that it is possible to estimate from some eccle-
siastical examples a growth of die cultivated area: in Catalonia, from 950 to
1000 some estates saw a growth of new cultivated lands amounting to 1 5% to
35% of the whole; similar figures have been proposed for Provence and central
Italy. Finally, we have the fundamental question: what was the volume pro-
duced by the cultivated area, whether or not diis increased by a third? We know
that the estimates made for the Carolingian period are appalling: seed pro-
duced twice or at best three times its volume, even if we ignore evidence sug-
gesting a weight for weight return, which would be an absurd negation of
agriculture. The few bits of evidence from the mid-tenth century at Brescia or
in the Maconnais suggest a ratio of 3-3.5 to 1, a very modest improvement.
But the other side of the millennium at Cluny we have arrived at 4-4. 5 . The 1 5
to 1 of Flanders in 1300 is still a long way off, but all progress has to make a
start.

The reader will perhaps have been surprised to have heard only about grains.
The reason is that about the rest, the cortrpanaticum (etymologically speaking ‘the
accompaniment’) we know nothing: at die foot of the Italian rocca there were
‘herbs and vegetables’ in die viridaria and orticelli , as also in Languedoc, perhaps
after the millennium. Elsewhere diere is silence about gathering berries, rabbit
warrens, and the eggs of the farmyard. Essential perhaps, if the weather
betrayed the peasant, but outside our reach. What about the vine, the source of
the Eucharist, the glory of the peasant, the honour to the table, the tradition of

Cambridge Histories Online © Cambridge University Press, 2008

ROBERT FOSSIER

antiquity? Enthusiastic historians have talked of an ‘explosion of viticulture’;
they have noted the southern European contracts of shared cultivation
between vines and olives, the drinking-bouts of castles and peasant commu-
nities, and stressed die generous rations allowed wine drinkers. But before
about 1100-25 it is n °t possible to talk about grape varieties, viticulture, wine
trade or quality.

Having stressed the importance of the part of the land reserved to animal-
rearing, not to mention the role which must be ascribed to hunting and fishing,
we find ourselves here still more deprived of reliable information. It has been
said that the pig was the animal to which most attention was given because it
was the basic source of meat, and this has been deduced from the practice,
already found in Carolingian times, of measuring the extent of woodlands by
the numbers of swine supported by them or capable of being supported by
diem, giving an approximate ratio of 0.75 hectares per pig. It is true that we do
not know whether the animals were really there, except for the rubbish-tips of
northern Germany mentioned earlier, which appear to suggest that cattle were
more important. We are reduced to general, supposedly common-sense con-
siderations, which are based on the taxes levied on acorns in the clearings of
the Weald (dens), or on the passage of transhumant flocks in the Pyrenees and
probably the Alps, though the first substantial flocks, in Italy for instance, date
from 1050 and later. Quarrels about woodland use, lawsuits over the fisheries
on the Saone, references to fisher- villages in the Fens of East Anglia or on the
Frisian terpen-, these are a poor soil where the historian can glean only die first
signs of a growth which was still taking shape.

SILVER ARRIVES IN THE COUNTRYSIDE

In an economy of waste, at best of gift and counter-gift, money as a sign of
exchange was unable to prevail. Exchanges were of favours, daughters, wine,
horses. Towns put up a short-lived pursuit of the coining of money and the
sale of luxury goods which can dazzle the Carolingian historian; but there is
none of this in the countryside. It was precisely the slow introduction of coins
or ingots into the rural world that was a powerful novelty in a growing
economy: die pump of exchanges between town and countryside sputtered
into life, and it would become an essential motor for the centuries following.

We have some difficulty in following the routes taken by silver from the
mines of Germany, Bohemia, or other less rich regions like western France or
northern Italy. Was there regular extraction, whether controlled by princes or
not, or was it more a case of liquifying thesaurised metal, accumulated espe-
cially by the church? Whatever the source, there was abundant striking of
coins: fEthelred II had 1 20,000 pounds of coins circulated; the mint at Pavia

Cambridge Histories Online © Cambridge University Press, 2008

Rural economy and country life

emitted 100,000 coins in 1020; In Catalonia, the parias imposed on Muslims
from 1018-35 onwards allowed coinages, gold in this case, on a very regular
basis. The number of mints rose sharply: more than twenty have been counted
for Picardy towards 1000, ten in Flanders, fifteen on the Meuse, if we confine
ourselves to considering northern France. Hoards recovered from this period
show a substantial quantity of denarii (pennies) in circulation: die Fecamp
hoard, abandoned around 985, contained 130,000 pieces.

Needs increased, in the town in particular, which is not our concern here.
But one should also take account of heavy and unavoidable expenses: between
980 and 1010 the Danegeld paid by die English to Scandinavia amounted to
1 50,000 pounds, and on a more modest scale the erection of a fortified tower
cost 2000, that of a mill 500. In order to build a church in the Boulonnais in
1017 the lords of the area had to sell a wood, two granaries and four mills. To
cope with these demands the lords were certainly able to count on the income
from commuting labour services, the extension of money rents, the expansion
of the taille; but in order for these furdier demands to be met there was a need
for peasants to have pieces of silver which could be screwed out of diem.
Where could these have come from if not from the sale of surplus foodstuffs
or craft products, or from a supplementary income? Between 975 and 1000 in
Catalonia die documented transactions conducted in silver coin amounted to
32% of those concerning foodstuffs, and 41% of those concerning catde and
horses, though only 15% of diose concerning manufactured products. At
Farfa in Central Italy in the same period almost all the renders were converted
to renders of coin.

Our documents are not distributed sufficiently equally to permit a geogra-
phy of the penetration of silver into the countryside. We only have a few hints
at a chronology: 945-75 on the coastlines of Catalonia and Languedoc, 960—90
in Italy and Aquitaine, not until the millennium and beyond north of the Loire.
In nordiern France and the Rhineland payments by weight or in heads of cattle
survived a long time, up to 1030 or 1050. But these were hangovers; by these
dates silver had already begun its role of economic and indeed social differenti-
ation within the village; a tripling of tire price of livestock has been estimated
for Spain and Italy between 975 and 1030, and at this last date a third of all
those who made wills in Catalonia had debts.

These were the timid beginnings of a silver-based economy. The founda-
tions of society were still land and freedom, and it was family ties, oaths and
rituals which kept them in place. The idea of a society without silver where
God had established a division of his creation into ‘orders’, each with its own
responsibilities, still remained the rule. In 1020 Adalbero of Laon was able to
express it forcibly, and the poet of Garin le Lorrain affirmed that ‘that which
makes for riches is not ornaments and treasures, but friends, for a man’s heart is

Cambridge Histories Online © Cambridge University Press, 2008

6z ROBERT FOSSIER

worth all the gold in the world’. 12 Was this an opinion still held, or merely nos-
talgia for a world which was disappearing?

We must conclude, and I shall do so in two ways, first by setting out what is
known about demography around the millennium, and then by summing up.
The question of population has been deliberately avoided up to now; to have
inverted the order of exposition by talking first about growth would have sug-
gested that this was undoubtedly a prior cause. I think that it was rather an
effect of the transformations which have been surveyed here, or, if one
prefers, a coincidental phenomenon, for, as we shall see, the dates where
expansion can be noted seem to be rather later than those of the developments
we have been discussing, though it must be conceded that research on this is
difficult. We have only two approaches: the study of cemeteries, though as has
been noted these were in the course of shifting at die time, allows us to say
somediing about the state of health and about the age of those who were
buried there; and lists of tenants or those owing labour-services kept by the
church, but here numerous and well-spread, in England (Evesham, Badi, Bury
St Edmunds), Germany (Fulda, Ghent, Gorze), Italy (Subiaco, Farfa), Spain
(Urgel, Braga). Coupled with the signs of increased exploitation of the land
revealed by pollen analysis, these data allow some quite precise observations.

The essential common feature is the beginning of demographic growth. It
was to last for three centuries. Can we date the beginnings? It was in 930-50 in
Sabina and Lombardy, 940-90 in Catalonia, 980— 1010 in Languedoc, Provence,
Poitou and die Auvergne, 1010—30 in Flanders and Picardy, Bavaria and
Franconia, Burgundy and Normandy, 1050-80 in England and the Rhineland,
after 1100 in central Germany. Attempts have been made to measure it
between its first signs around 950 and die mid-eleventh century, a period which
represents the first phase. One author has estimated the global figure for
European population rising from 42 to 46 million inhabitants; another con-
fines himself to a rise from 20 to 23 million. These suggestions are interesting,
but lack any kind of proof, though one can accept the estimation of a slow
annual rise in western Europe, amounting to 11% in die first half of the
eleventh century, modest but regular, or die figures for die average number of
children born to a fertile marriage, rising from 3.5-4 to 4-5.3 between 980 and
1 o 5 o. It goes without saying diat these figures cannot do more than show a ten-
dency, for too many data escape us. Was it an improvement in nutrition which
caused a decline in mortality? Swedish and Polish cemeteries of the millennium
still contain 20-30% of children aged less than five. Or was it the ‘hidden
infanticide’ practised against daughters by giving them less care that declined?
Or was the social change of earlier marriage accompanied by the physiological

12 U rowans de Garin leLoherain u, verse 268 .

Cambridge Histories Online © Cambridge University Press, 2008

Rural economy and country life

change of a growth in wet-nursing, creating a tendency in favour of births? All
these problems confront the demographer, whose only certainty is that there
were more and more people.

A few remarks will suffice to sum up. The climate may have been better,
there were certainly more people, the family was set on a new basis, the frame-
work of the village was stable, die seigneurie with its guarantees and restriction
was being put in place: this is die balance-sheet of the decades around die mil-
lennium. What about the ‘terrors’ invoked by die romantics? In 1000, as in
1033, people may have thought about the birth or death of Christ, but they had
enough to do to make a living; there was no need to worry about dying. On the
contrary, they were participating in a ‘birth’, that of Europe, and they were
conscious of it. How odierwise can we conclude dian by citing the words of a
Burgundian monk and a German bishop: ‘The world, shaking off the dust of
its senility, seemed to cover itself everywhere with a white robe of churches’, 13
and ‘at the thousandth year after the birth of Christ a radiant dawn broke over
the world’. 14

13 Radulf Glaber, Historiae iv, 5 . 14 Thietmar, Chronicon vi, 1 .

Cambridge Histories Online © Cambridge University Press, 2008

CHAPTER 3

MERCHANTS, MARKETS AND TOWNS

Peter Johanek

the beginnings of the European town in the form known to us from the
late middle ages lie in die tenth century. Urbanism began its dynamic phase in
the late eleventh century, reaching its climax in die thirteenth, but the basic ele-
ments were assembled between the decomposition of the Frankish empire at
the end of the ninth century and the early decades of the eleventh. In this tran-
sitional period the commercial revolution began.

The renewed rise of the town as a social formation is certainly closely con-
nected with the extension and intensification of trade: merchants are therefore
an important group in the shaping of the medieval town, in its topography, its
institutions and its social networks. Their activities were the most spectacular
and impressive, and occasionally overshadow the contribution and activities of
the other forces driving developments forward.

The rise in urban development and the changes in die structure and organ-
isation of trade which will be described here presuppose a general expansion
of the economy and an increase in prosperity, especially in the agrarian sector.
This is die only explanation for die emergence of a broad stratum of consu-
mers able to absorb the goods brought by long-distance trade. From the tenth
century onwards this stratum was multi-layered, from clerics and aristocrats
acquiring rich oriental cloth to wrap relics in, down to the Frisian manorial offi-
cials of the monastery of Werden on the Ruhr, who in the elevendi century had
to make renders of pepper and wine to their clerical lords. And indeed the
whole of Europe, including the Byzantine empire, shows an evident rise in
agrarian production and demographic growth, though obviously there were
variations between individual regions.

Those tenth-century Europeans who drew maps of the world did so com-
pletely in die tradition inherited from antiquity: they stressed Europe, and in
particular the Mediterranean, which was presented as the centre of the conti-
nent, from which its other parts and the world outside were viewed. In reality
the Mediterranean was a meeting point between die Islamic and die Christian
worlds, Christian meaning here both Greeks and Latins.

Cambridge Histories Online © Cambridge University Press, 2008

i Urban settlements and emporia in the Scandinavian and Baltic regions

Cambridge Histories Online © Cambridge University Press, 2008

PETER JOHANEK

A Mediterranean observer, especially one from die Islamic regions, looking
at Europe, would have been confronted by three different regions of monetary
circulation. In Islamic Africa and in Syria, as in Byzantium, gold dominated,
but there were other coinages of silver and copper. Whereas in Byzantium the
nomisma (bezant) entered a crisis at die beginning of die tenth century, Islamic
North Africa was able to acquire new gold bullion from sub-Saharan Africa. In
general this region was characterised by a highly differentiated monetary
system, though in Byzantium this was more concerned with the fiscal needs of
the state than with trade. Alongside this south-eastern region we find
Carolingian Europe, with a monometallic silver coinage and a close connection
between markets and mints. This region included die Anglo-Saxon kingdoms
and Islamic Spain, which had already made die transition from gold to silver in
the eighth century. Finally, in the north, around the Baltic with its Slav and
Scandinavian coastlines and their hinterlands, we also find noble metals used
for payment. But die hoards on which our knowledge is based include hack-
silver as well as coins, which suggests that it was not coins but metal measured
by weight which served as a medium of exchange. Up to and beyond the mid-
tenth century, to around 960 or 970, these hoards were dominated by Arabian
silver coins from Transoxania, whose mints were fed by the local silver mines.
The coins penetrated beyond the Baltic into the Reich, for die oriental traveller
Ibrahim ibn Ya'qub saw in either 961 or 965/6 dirhams from die mint at
Samarqand in Mainz. 1 These dirhams subsequendy disappeared, and the Baltic
was then dominated by pennies from German and Anglo-Saxon mints. This
change was certainly brought about by die discovery of additional silver
deposits in the Harz (especially at Rammelsberg near Goslar), but it was also
the result of the enhanced economic power and die active trade of Ottoman
Germany and Anglo-Saxon England.

Our hypothetical Mediterranean observer would thus have perceived a grad-
uation in forms of trade and exchange of goods. North of the Alps and the
Pyrenees, and especially in the region around the Baltic, these forms were
simpler and less differentiated, but even here the use of coined metal inten-
sified in die course of the tenth century. Nevertheless, the Mediterranean
remained the real region of urban culture in the tenth century. This culture
rested in part on ancient tradition, but it also developed a powerful dynamic of
its own. The Islamic regions, from Mediterranean Spain through to Egypt and
Mesopotamia, were noticeably different from the European economy, with
which they had intensive contacts. Here we find really large and economically
active towns, which can be matched in both eastern and western Christian
Europe only by Constantinople. The trading of Islamic merchants was shaped

1 A rabische Berichte , p. 3 1 .

Cambridge Histories Online © Cambridge University Press, 2008

Merchants, markets and towns

by a detailed legislative framework based on writing. Communications between
merchants, information about profit and loss, about the availability of goods
and means of transport, and about delivery dates were also as a rule carried out
in writing.

This urban culture of Islam, which may be seen as a religion of merchants,
can be set alongside very varied forms of urbanism in Christian Europe and
the pagan north. In Italy there was a great continuity of urban life from Roman
times; the civitates had remained centres of secular and ecclesiastical administra-
tion and nodes of long-distance trade-routes, even if their architectural land-
scapes had been fundamentally altered by changes in the practice of patronage
in the erection of public buildings. North of the Alps, in Gaul and the formerly
Roman parts of Germania, most of the civitates had shrunk considerably, often
being reduced to a core area which functioned as citadel or fortification. In the
regions outside the old Roman empire incorporated into the Frankish empire
during the Merovingian and Carolingian periods, and especially in the Slav and
Scandinavian regions and in the British Isles, we find very varying beginnings
for quasi-urban settlements and for mercantile centres. It must be emphasised
that at the beginning of the tenth century trade and crafts were not inherently
bound up with die social form of die town in these regions, often being organ-
ised in connection with lordship outside civitates , especially in the lordships of
the great monasteries. Writing was used on a large scale in the organisation and
regulation of trade only in Byzantium; outside the Mediterranean region law
was confined to symbolic forms for concluding contracts. Lay literacy, which
survived to some extent in Italy, evidently declined sharply in the course of the
tenth century. In the Scandinavian north we find an increase in runic inscrip-
tions, especially around trading centres, but there are no signs of a rune-based
mercantile literacy as known from the twelfth century onwards through
archaeological finds from Bergen. Only in Haithabu has a runic staff been
found, datable to about 900, which may perhaps be interpreted as a merchant’s
letter.

In spite of this, Europe showed itself an attractive trading partner for the
urban culture of Islamic north Africa and the N ear East, and indeed it was pre-
cisely these trading links which lay behind the flowering of Islamic trading
centres on the southern Mediterranean coast in the tenth and eleventh centu-
ries. The cities of the Arab west, especially in the Maghreb and in al-Andalus
(Spain), formed the far end of a chain of cities linking tire Mediterranean with
the Indian Ocean in a unitary trading zone in which goods from Asia, espe-
cially spices and luxury goods, flowed to northern Africa and Europe. The
Arab west not only expected the same self-evident standard of luxury found in
Damascus and Baghdad, it also, through its contacts with west African gold
production, disposed of considerable economic strength. In addition, the

Cambridge Histories Online © Cambridge University Press, 2008

PETER JOHANEK

tenth century saw the height of Islamic power in die western Mediterranean,
even if Islamic unity had disintegrated. The Shi’ite Fatimid dynasty established
itself in 909 in Kairuan (Ifrlqiyyah) and in 969 it conquered Egypt; the
Ummayad amirs of al-Andalus took the tide of caliph in 912. This all lent
added weight to the region, whose large cities, especially Cordoba and al-Fustat
Cairo, developed rapidly.

Cordoba, the seat of the Ummayad caliphs, grew sharply in the ninth and
especially in die tenth centuries: estimates of its population in the period vary
from 90,000 to 500,000, even a million, though the first figure is more realistic.
The town was an agglomeration of different settlements, owing their origins to
the rulers’ initiative. Besides die old city (Medina) widi a palace and a central
mosque there were other palace cities in the immediate vicinity: al-Rusafa and
Madlnat al-Zahra (Cordoba la Vieja) under c Abd al-Rahman III (912-61) and
al-Madlnat al-Zahirah under the dictator al-Mansur (around 980). The length
of the walls around ancient Medina was only 4 km, but at the beginning of the
eleventh century there was a moat of some 22 km around the agglomeration,
and the palace city al-Zahra remained outside this. Cordoba was one of the
places at which the west encountered Islamic urban culture: Abbot John of
Gorze stayed here between 95 3 and 956 as ambassador of Otto I, guided by
merchants from Verdun familiar with the country. Their impression is reflected
in the phrase used by Hrotsvitha of Gandersheim to characterise the city: decus
orbis, the ornament of the globe. Cordoba combined trade, specialised crafts
for the production of luxury goods (especially leather), and administrative and
military functions, with a strong garrison. It was also a centre of learning with
an extensive book production, evidently also organised as an industry. Above
all it may be seen as an exceptionally large centre of consumption, functioning as
tire metropolis for an economic region comprising Spain and western north
Africa around Fez, and managing its marine trade from Almeria, the port
founded by ‘Abd al-Rahman III in 9 5 5 .

Comparable with Cordoba in its character as an urban agglomeration
created by the ruler was Kairuan, established like Cordoba away from the
coast, and the starting point of Fatimid expansion. But the real pendant in
north Africa to Cordoba was al-Fustat on the Nile, immediately south of the
city of Cairo founded by the Fatimids in 969. It grew together with Cairo into a
single city, but remained the economically dominant part until well into the
twelfth century. Founded in 642 as a garrison by the conquering Arabs, it had
developed by tire tenth century into a huge agglomeration of separate quarters
(about twenty all told, with an average size of 20-40 hectares; al-Qarafa had
300 hectares). Each of these was assigned to a tribe from the conquest era and
organised according to its laws. The Arab geographer Ibn Hawqal (d. 988) esti-
mated that al-Fustat was about a third of the size of Baghdad, and the popula-

Cambridge Histories Online © Cambridge University Press, 2008

Merchants, markets and towns 69

don in 969 was probably somewhat under 100,000. From then it grew rapidly
and in the eleventh century it lay somewhere between 300,000 and 500,000.

This growth was quite evidendy based on the extraordinary economic pros-
perity about which the encyclopedist Mas ‘udi (d. 956 or 957) reported: ‘All the
kingdoms located on the two seas which border the country bring to this com-
mercial centre all the most remarkable, the rarest, and best perfumes, spices,
drugs, jewels and slaves, as well as staples of food and drink, and cloth of all
sorts. The merchandise of the entire universe tiows to this market.’ 2 The deci-
sive push came around 1000, when the seizure of power by the Karmates in
Bahrein made sea transport in the Persian Gulf so dangerous that the great
bulk of trade from the Indian Ocean to the west henceforth came over die Red
Sea via Aden, ‘Adhab on the Sudanese coast and Qusan on the upper Nile to
Egypt, thence to al-Fustet and Cairo. The cities of Syria and to a lesser extent
Byzantium were still the final destinations of the caravans, but al-Fustat and its
port of Alexandria became the most important emporia of the Mediterranean
region.

This is true above all for exchanges between the Islamic world and Christian
Europe, which were also stimulated by the military needs of the Fatimid
dynasty, whose demand for iron and wood for ship-building could be met only
by imports from Italy. There were also traditional imports of goods from die
Occident, listed already in the ninth century by the geographer Ibn
Khordadhbeh: slaves, furs of all kinds, and swords. 3 They can be seen in the
presents made by the margravine Bertha of Tuscany in 906 to the caliph al-
Muktafi: swords and male and female slaves from the Slav regions. In 949
Liudprand of Cremona also brought weapons and slaves to Byzantium when
acting as ambassador for Berengar II. 4 For slaves especially there was an
extraordinary demand in the Islamic lands, and indeed in Spain the whole
system of government was largely based on slaves from Sclavinia. John of
Gorze was accompanied by merchants from Verdun on his mission, and
Liudprand of Cremona reports that the Verdunese merchants had become
particularly rich by trading in eunuchs with Spain. 5 This flow of trade to the
Islamic Mediterranean thus reached deep into Christian Europe, as far as tire
east Frankish realm, and the rise of Liudolfing Saxony in the late ninth century
may be due among other tilings to the fact that it was the source of these wares.

Islamic merchants did not extend their activity beyond the boundaries of
Islamic rule, nor did Islamic rulers encourage activities of this kind. Rather,
they allowed foreigners into their own territories to trade with them, though
they did not allow transit passages. In Cordoba and in the rest of Islamic Spain

2 Cf. Staffa (1977), p. 46. 3 Kitab alMasdlik wal-Mamalik , p. 1 14.

4 Cf. Gil (1974), pp. 3 1 o— 1 1 ; Liudprand, A.ntapodosis vi, 6, pp. 1 5 5—6.

5 John of Saint- Arnulf, Vita Iohannis abbatis Gor^iensis, c. ii7,pp. 370— 1; Liudprand, A.ntapodosis \i, 6.

Cambridge Histories Online © Cambridge University Press, 2008

7 °

PETER JOHANEK

these traders came from the Frankish realms, while al-Fustatwas visited above
all by Italian merchants from Amalfi, as we shall see. But die most important
group in these exchanges was Jewish merchants. They played a leading role in
inner-Islamic long-distance trade, as is shown by the documents from the
Geniza of Cairo, which begin towards the end of die tenth century. They were
not a substantial part of the population, especially considering the population
figures named for the large cities. In eleventh-century Egypt there were prob-
ably no more than 1 5 ,000 of them, and their most important centre was
Alexandria, not al-Fustat. But Ibn Khordadhbeh speaks in his report on
western trading goods of Radhanites, Jews who were based in die Christian
west, probably in southern France, and who carried out a far- dung trade as far
as India and China. 6 Jews did indeed play a leading role in the long-distance
trade of the Frankish empire from the ninth century onwards, favoured by the
privileges granted by Louis the Pious. They were setded here, owning land,
vineyards and mills, above all in southern France, for example in Narbonne
where they are mentioned in 899 and 919, in Saintes (961) and in Vienne
(975-993), but also in Regensburg, where in 981 the Jew Samuel sold a rural
estate to the monastery of St Emmeram. 7 Their scattered communities were
concentrated along important trading-routes, especially the Rhine. The refer-
ences in charter sources show diat diey were seen as long-distance traders par
excellence. The Raffelstetten trading ordinance (903-906), which regulated the
salt trade along the Austrian Danube, calls them ‘the merchants, that is the Jews
and the odier merchants’. 8 Similar phrases are used in privileges for Magdeburg
of 965 and 979 and for Treviso of 991, while in Byzantium die Book of the
Eparch, the main source for the trading history of Constantinople in the tenth
century, uses die phrase ‘Jews or merchants’. 9

The activities of these Jewish merchants evidently encompassed the whole
of continental Christian Europe, extending into Sclavinia and perhaps into
Scandinavia, as is suggested by die fragments which can be deduced of a travel
report by Ibrahim ibn Ya'qub, a Jew from Spain, for he describes Prague as a
slave market and Haithabu as a heathen trading-centre with only a few
Christians. 10 On the other hand we can see die continuous links to the Islamic
regions and their economic centres, even if not all Jews setded in the
Carolingian successor-states undertook such long journeys as the Radhanites.
Jewish mercantile activity, which reached a marked peak in die tenth and
eleventh centuries, linked the Islamic world with Europe and filled die conti-

6 Kitab al-Masalik wa’l-Mamdlik , pp. 1 14— 1 5 .

7 DD Ch S 23 and 102; Lot (1950), pp. 540—1; Cartulaire de l' abb aye de Saint-A.ndre-le-Bas de Vienne , no. 91,
p. 68 (cf. Endemann (1964), pp. 130— 1); D O II 247.

8 MGH Cap., no. 25 3, II, p. 252.

9 D O I 300; DO II i98;DO III 69; Le Livre du Prefet , p. 3 3 . 10 A.rabische Bericbte , p. 29.

Cambridge Histories Online © Cambridge University Press, 2008

Merchants, markets and towns 7 1

nent with oriental mercantile culture. The comparatively richly transmitted
Responsa literature of the ninth to the eleventh centuries, in which Jewish mer-
chants posed questions about trading law to legal experts, shows just how far
this trading culture was governed by literacy. The statement made by one of
these experts in the eleventh century is valid for these traders: ‘they used to
conduct their affairs by letters which they wrote to one another. And it was
their practice that . . . letters were as binding as their words.’ 11

Islamic urban culture thus influenced Christian Europe through Jewish mer-
chants, but these exported goods alone, not the urban forms and institutions
of Islamic cities. Here there was no exchange, not even as a result of the expe-
riences of European traders in Islamic lands.

The Islamic cities were centres of dynastic and religious power, controlled
by the c umma, the Islamic state community. The administration of these cities
was — even though Islam was a mercantile civilisation right from the beginning
and merchants enjoyed a high social prestige - run by officials of the ruler and
his agents. There was no special community of self-administering citizens;
only tire non-Islamic segments of the population (Jews and Christians above
all) enjoyed a certain autonomy. The absence of a community of citizens and
the social fragmentation of the Islamic city into ethnic, religious and profes-
sional groups also affected its topography. Normally there was no regular
network of streets linking all die parts of the city; rather, we find an agglomer-
ation of quarters complete in themselves. The main features of their topogra-
phy were the palace, the Friday mosque and school as religious centre, and
above all the inner-city market, normally situated next to die mosque, which
offered the products of urban crafts. Markets for wholesale and long-distance
trade, merchants’ inns, and also markets for the agrarian produce of the hinter-
land lay on the periphery. The typical Arab praise of the city stresses besides
palaces, mosques, the learning of the schools and the abundance of the
markets, the gardens and baths, and above all the number and size of the
houses. Ibn Hawqal notes proudly diat al-Fustat’s and Cairo’s houses had five,
six or even seven storeys and the Persian traveller Nasir-i-Khusrau had the
feeling of having a mountain before him when contemplating Cairo around
the year 1000. 12 Admittedly, not all Islamic cities reached the size of Cordoba,
Kairuan and al-Fustat, but in the Christian lands there was only one city which
could be compared with these Islamic metropolises of the Mediterranean
region: Constantinople. From this city 7 there also ran in the tenth century an
important trading-route for the import of oriental wares into Europe, a trade
in which those same Italian cities whose merchants traded with al-Fustat parti-
cipated.

11 Cf. Ben-Sasson (1976), p. 398. 12 Cf. Wiet (1964), pp. 36, 39— 40.

Cambridge Histories Online © Cambridge University Press, 2008

7 2

PETER JOHANEK

The Byzantine empire, at the height of its medieval power in the tenth
century, was a little smaller than the western empire, but it possessed a genuine
capital, tire largest city in Europe at tire time. The walls enclosed an area of 24
km 2 , and estimates of its population extend to a million, though
250,000-300,000 is probably nearer tire truth. Constantinople was the heart of
an empire with a strongly centralised provincial administration, which came
increasingly under the control of the metropolitan elite just at this time. The
numerous cities of the Byzantine provinces, in particular those of the Balkan
peninsula, came nowhere near the metropolis in size. Even Thessalonika, the
most important city after Constantinople, had an area of only 3.5 km 2 , and
most of these cities were presumably very small. They were also not pre-
eminendy centres of craft and trade, least of all long-distance trade, but rather
centres of consumption orientated towards their hinterlands, where rich land-
lords, following the tradition of Roman antiquity, consumed tire surplus wealth
of agrarian production. No class of economically active burghers developed
here.

To some extent this statement is valid of Constantinople itself. The Book of
the Eparch, a collection of laws probably published by Leo VI around 9 1 1-1 2,
names a great number of crafts and groups of merchants, 13 but these were very
strongly aligned with the needs, of metropolitan consumers with a high stan-
dard of living. Here too the rich landowners dominated alongside imperial offi-
cials. The Byzantine economy, in spite of the increased prosperity of the
provinces in the tenth century, tended towards autarky, by contrast with the
Islamic world. The Book of the Eparch itself shows that there was a considerable
import of oriental wares, especially from Persian regions, and also a native pro-
duction of luxury goods (silk-production, purple-dyeing), but we can hardly
discern long-distance trade and brokerage aiming beyond tire borders of the
empire. Greek merchants, who had carried out a good deal of early medieval
European long-distance trade, from the beginning of the tenth century no
longer went abroad: the last reports of Greek merchants in southern France,
for example, date from 9 2 1 . 14

The Byzantine empire had traditionally organised its trading contacts, both
with western merchants and with the Russian and Islamic regions, at fixed
points on the border. The admission of foreign merchants to die capital itself,
a procedure whose forms become visible in the tenth century, came to be of
great importance. But the Book of the Eparch imposed a strict regimentation on
native crafts and trade, and the same happened to foreign merchants. They had
decisive restrictions placed on the length of their stays (as a rule three months);
they had fixed living-quarters ( mitatd ) in which they could be strictly controlled;

13 Le Livre du Prefet, passim. 14 Recueil des actes des rois de Provence, ed. R. Poupardin, no. 39 , p. 108 .

Cambridge Histories Online © Cambridge University Press, 2008

Merchants, markets and towns

and for particular wares there were export prohibitions or restrictions on the
quantities which could be exported, especially in wares of the highest quality,
which were retained for the Byzantines’ own needs. Liudprand of Cremona
experienced all this when, on his departure from Constantinople in 968, five
pieces of purple cloth were confiscated by customs officials; it was in vain that
he protested that the merchants of Venice and Amalfi were able to export such
textiles from Byzantium and offer them for sale in Italy. 15

In the course of tire tenth century, the contacts of the Italian cities with
Byzantium and tire Islamic world do seem to have intensified in spite of the
restrictions found in normative sources. Amalfi had the greatest successes, but
Venice the most lasting ones; Pisa and Genoa appeared on the scene only
around the millennium. Both Venice and Amalfi had their roots in Byzantine
rule over Italy, and this alone orientated them from the start towards the
Levant trade. Amalfi was one of the castra erected by the Byzantines against the
advancing Lombards towards tire end of the sixth century. Almost inaccessible
from the land, built on a tiny territory, but endowed with an excellent harbour,
it began its rise in the ninth century, especially after it had freed itself from sub-
jection to Naples in 840; like the latter city it pursued, though independently, a
policy of occasional cooperation with the Arabs. This led to an early link with
north Africa, with the Aghlabids and later the Fatimids in Kairuan and their
harbour al-Mahdiyyah, recorded from 870 at the latest. It is therefore not sur-
prising that, following the Fatimid conquest of Egypt, al-Fustat/Cairo
belonged to their destinations. One hundred and sixty Amalfitans, ‘who had
come there with their wares’, perished in a pogrom in 996. 16 This suggests a
real colony, encouraged by the on the whole xenophile policy pursued by the
Fatimid rulers. The Amalfitans pursued a triangular business. They brought
corn, linen, wood above all and perhaps iron in exchange for gold and spices to
Tunisia and Egypt. The gold paid for the imports of textiles, jewels and other
luxury items from Byzantium. These activities intensified towards the millen-
nium, and tire Amalfitans perhaps received permission to settle in Antioch and
Jerusalem around that time; they had been resident in Byzantium from the
beginning of the tenth century and backed Constantine VII in the rulership
crisis of 944. Because of its Arab connections, Amalfi was probably the most
important Christian trading centre in die Mediterranean around the year 1000,
ahead of all other south Italian cities but also of its rival Venice.

Ultimately, however, Venice was more successful. This city too, which had its
origins in a settlement established by refugees who had retreated before the
Lombards around 600 to the islands of the laguna, had been important since
the early nindi century. It profited from its special political position, which

15 Liudprand, Relatio, c. 5 5 , p. 205 . 16 Cahen, ‘Un texte peu connu’.

Cambridge Histories Online © Cambridge University Press, 2008

PETER JOHANEK

allowed it to appear as a member of the Byzantine empire and so gave its mer-
chants access to Constantinople; from 880 on it was independent of the
Frankish empire. Venetians, like Amalfitans, traded with the Arabs, in part in
the same militarily significant goods, and this led to friction with Byzantium in
97 1 . But in 99 2 the Venetians secured a treaty which gave them primacy within
Constantinople; in 1082 this culminated in a monopoly, while Amalfi became a
backwater following the Norman conquest of 1077. These political facts were
important, but Venice also had advantages which Amalfi could not offer: it was
in a position to provide die Levant trade with a large-scale and receptive hinter-
land, northern Italy, which was rich in civitates and economically active, and also
to open up the transalpine trade. The caput Adriae between Istria and die mouth
of the Po, along with the Rhone valley, had always been the main entry points
for Mediterranean wares into central and nordiern Europe. Venice was able to
bring diis region and in addition a part of the eastern Adriatic coast under
political control. Its rival Comacchio was eliminated in 9 3 3 , but it is noteworthy
that Venetian activities to secure intiuence over the harbours of the northern
Adriatic intensified around the millennium, when Venice was cooperating
more intensively with Byzantium.

The securing of Ottoman rule in Italy placed Venice’s access to transalpine
regions on a firm political footing and hence facilitated it. In 967 begins the
long series of Ottoman pacta widi the city, based on ancient tradition: diese
granted the Venetians freedom of movement in northern Italy, especially for
the trade with the most important cities between the Adige and the Po. 17 For its
trade Venice thus disposed of two privileged zones and enough political influ-
ence to be able to restrain all potential rivals in die region at the head of the
Adriatic. This made it in the long run the most important interface between the
Levant trade and die wares of transalpine regions, which in the tenth century
certainly included slaves and furs, perhaps also metals. Venetian trade was evi-
dendy closely observed in Germany: already in 860 it was known in Fulda that
goods flowed into Venice, and Thietmar of Merseburg noted in his Chronicon
under 1017 that four Venetian ships widi all kinds of different spices had
suffered shipwreck. 18

In the tenth and elevendi centuries German traders were not yet to be found
in Venice. Foreigners used the city merely as a starting point for journeys to
Byzantium, and Venetian ships as a means of transport, like the ‘very rich mer-
chant’ Liutfrid from Mainz, whom Liudprand of Cremona met serving as an
ambassador of Otto I’s in Constantinople in 949. 19 Exchanges between
German merchants and Venetians were evidendy carried out in Treviso 20 at the

17 Cf. Rdsch (1982), pp. 7—8. 18 Annales Fuldenses, s.a. 860, p. 54; Thietmar, Chronicon vii, 76, p. 492.

19 Liudprand, Antapodosis vi, 4, p. 154.

20 Cessi (ed.), Document i relativi alia storia di Venecia, no. 1 89, pp. 1 82—4; cf. Rosch (1982), pp. 80—1 .

Cambridge Histories Online © Cambridge University Press, 2008

Merchants, markets and towns

foot of the Alps, where at the beginning of the eleventh century a German
toll-station ( ripaticum teutonicorum ) is recorded, and above all in Pavia.

Pavia had been die capital of the Lombard kingdom, and in Ottoman times
also it was one of the preferred sedes of the rulers when they were in Italy,
alongside Ravenna and Rome. The central administrative apparatus of the
regnum Italiae with its base in the royal palace at Pavia apparently remained
intact. From shortly after the death of Emperor Henry II in 1024 there sur-
vives a list of die revenues of the royal chamber, the Instituta regalia, also known
as the Honoratiae civitatis Papie , which shows Pavia at the centre of long-distance
trade in northern Italy. 21 It directs attention on the one hand to the ten trading
stations in all, the clusae , situated at die entry to the Alpine passes from Susa in
the west to Cividale in die east, and to die merchants coming from die north
(among whom numerous Anglo-Saxons evidendy enjoyed a privileged posi-
tion) and dieir wares: horses, slaves, wool and linen cloth, tin and swords. On
the odier hand we find the Venetians and merchants from the south Italian
cities, Salerno, Gaeta and Amalfi, who brought in oriental and luxury goods:
spices, ivory, mirrors and valuable textiles. At the end of the ninth century
Notker of St Gallen had already described the great variety of textiles available
from the Venetians to Frankish magnates, and Odo of Cluny in the first half of
the tenth century relates how Count Gerald of Aurillac had been offered silk
and spices before the gates of the city. 22 Pavia thus appears as a market which
was frequented both by rich consumers themselves and by traders, a meeting-
and exchange-point between the region north of the Alps and the
Mediterranean. The city was a focus for this trade because it was the centre of
government in the regnum Italiae, not because of any potent stratum of mer-
chants of its own. The role of Pavia, rooted in older relationships, was revived
and intensified by the Ottomans’ policy in Italy. The increasing frequency of
trade on the Alpine route from the Rhine during the second half of the tenth
century is also visible in what was evidently the very rapid development of
Zurich, which lay in the northern hinterland of the Biindner group of Alpine
passes. This concentration of long-distance trade on Pavia and the city’s role as
a centre of distribution vanished after the royal administration had disinte-
grated in the course of Henry IPs reign and the Pavians destroyed the royal
palace.

This action directs our attention to the remaining cities of northern Italy,
whose inhabitants also began to develop their independence in the course of
the tenth century. This indicates a new stage in urban history. The revival of
long-distance trade, in particular the strengthening of communication

21 Cf. Die ‘Honoratitie Civitatis Papie’ , ed. Briihl and Yiolante, passim.

22 Notker the Stammerer, GestaKarolin , 17, p. 86; Odo of Cluny, T ita S. Geraldit, 27, col. 658.

Cambridge Histories Online © Cambridge University Press, 2008

PETER JOHANEK

between two great trading regions, as seen around 1000, gave a powerful shove
to economic development and to urbanism. But it coincided with a general
intensification of medium- and short-distance trade and a flourishing of hand-
icrafts. All these together favoured the development of urban and quasi-urban
forms of life and social organisation in varying degrees. All die post-
Carolingian kingdoms were affected by diis. Besides the growth of and the
emergence of specialisation widiin die civitates we find their penetration of
the hinterland with places for the exchange of goods taking the legal form of
the market ( mercatum ), which provided those who traded and also the produc-
ers of craft and agrarian goods with a stable framework for their activities:
peace and protection bodi at the market itself and while travelling to and from
it, legal security and the setdement of disputes arising out of transactions,
together with reliable monetary conditions.

This process was stimulated and encouraged by rulers and odier lords, who
guaranteed the legal setting and derived fiscal benefit from market dues, in par-
ticular from tolls. True, market foundations are not an innovation of the
Ottoman period, but go far back into the Carolingian era. Yet in die tenth and
eleventh centuries diey reached a new stage of development, and were used
deliberately to intensify lordship in the central regions within which the medie-
val town developed, that is in Italy, France and Germany.

In Italy the network of civitates was finer meshed than in the transalpine
regions, an inheritance from antiquity: the distances between episcopal sees
ranged between 15 and 50 km. In the transalpine regions they were much
greater; even in die German regions west of the Rhine and in Lotharingia
they were 50 to 130 km, and furdier east drey could be still larger. For this
reason no additional quasi-urban settlements developed alongside die episco-
pal sees in Italy: urban life is congruent widi die episcopal city. The civitates in
Italy evidendy suffered less in the course of the Germanic incursions of late
antiquity than the episcopal sees in Gaul, and they did not experience so
great a shrinkage. The walled area of the more important towns varied
between 20 and 40 hectares, and even the great exception of Rome, with an
area of 13.86 km 2 , did not achieve the extent of die Byzantine and Islamic
metropolises.

Rome played no active economic role. For several centuries it had been a
rural town widi large farmed areas widiin the walls. At best it was a centre of
consumption. The luxury goods brought by long-distance trade, above all drat
of die Amalfitans, flowed into the courts of the popes and their clergy, to the
numerous churches and their decorations, and to the crowds of pilgrims who
visited the tombs of the apostles. Evidently no long-distance trade was plied by
Roman merchants. Politically, both city and papacy were in die hands of rival
aristocratic families, and even Ottoman rule faced constant revolts: ‘Rome and

Cambridge Histories Online © Cambridge University Press, 2008

Merchants, markets and towns

the papacy were at their nadir.’ 23 Rome’s importance for urban history was not
a product of its political or economic role but of the picture of Rome as an
urban caput mundi, as the city of Christianity itself, shaped by tradition and
renewed by Otto III.

The centres of urban innovation in Italy lay in Tuscany and above all in
Lombardy. The basis for the civitated economic development was die fertility of
the Po basin and its tributary valleys. Liudprand of Cremona formulated this
almost epigrammatically when he said that Venetians and Amalfitans brought
valuable textiles to northern Italy in order to sustain their existence widi the
foodstuffs bought in return. 24 But undoubtedly the export of agrarian sur-
pluses was the main driving force behind the early rise of die Lombard cities
and the source of dieir prosperity. The landowning nobility of Italy, unlike diat
of Gaul, had never left the civitates , and so city and hinterland remained closely
linked. The civitas retained an oligarchic structure even within its walls.
Although the bishop, as elsewhere, was the most important figure in the city,
and his position was further strengthened by Ottoman privileges, he was still
not the real ruler of the city, but had continually to deal with odier groups of
the urban population and the distribution of power among them. The popula-
tion was subject to a unitary law, and consisted for the most part of the free. To
be able to defend the extensive ring- walls the population bore arms, and took
part in the conventus , a popular assembly. The nobility naturally played a decisive
role in this highly differentiated urban society, and die bishop and the other
officials in the city were in effect merely the exponents of the aristocracy and its
factions. But their election was the product of inner-urban decision-making.
The permanent market within the walls was the economic centre of die city,
already equipped with densely built-up market stands often owned by eccle-
siastical institutions. It was here that the activities of traders and of the urban
craftworkers intermeshed most closely. The differentiation in craftwork and its
concentration within die city seems to have been a very important factor in
determining the economic power of a city; it was almost as important as trade.

The significance of crafts can be seen in die rise of Milan, which around
1000 was probably already on a par with Pavia. It was not inherendy favoured
by its position away from the Po, but it was able to concentrate long-distance
trade on itself because its archbishop could guarantee the safety of traders
along the Alpine route to Chur. This underlines once more the importance of
this route for Italian trade with the transalpine region, whereas the western
Alpine passes still suffered from the depredations of Saracen bands, who in
891 had set up a base in Fraxinetum, between Marseilles and Nice. From here
they made razzias by water and land, and they were not driven out until 973.

23 Krautheimer (1983), p. 145. 24 Liudprand, Relatio, c. 5 5, p. 205.

Cambridge Histories Online © Cambridge University Press, 2008

7 «

PETER JOHANEK

That also helps to explain the delay before Pisa and Genoa, with their excellent
harbours, were able to take a leading part in long-distance trade. A decisive
contribution to Milan’s prosperity was made by the development of a produc-
tive ironworking industry. This profited from ore deposits on Lakes Maggiore
and Como, largely in the hands of the Milanese monastery of S. Ambrogio. At
all events it is smiths and ironworkers alongside merchants whom we find
among those Milanese citizens who acquired land in the surrounding regions
around 1000.

Processes like that just mentioned demonstrate the economic superiority of
the civitas , as does the fact that around 1000 the price for land in Milan was
thirty-six times as high as in die countryside. 25 But the countryside was also
subject to increased commercialisation, as seen from royal diplomata granting
rights of market, which begin before the middle of the century. Bishops pos-
sessed such rural markets, as did individual monasteries and nobles, such as the
Vuaremundus who received in 948 from King Lothar the right to collect all the
dues pertaining to the king on contracts concluded in his castles and villages or
in markets which he might erect in places belonging to him. 26 Trading and the
market are here linked widi castle-building, incastellamento, a practice whereby
nobles and ecclesiastics sought to intensify their lordship. These markets and
fortifications were only rarely the basis for urban formation. Urban develop-
ment generally remained linked with the civitates and the marketplaces within
them. They grew through the accumulation of burgi, unfortified settlements
outside the walls, which were incorporated into the civitas by the walls built in
later eras.

It is obvious that the merchants of die Italian civitates belonged to the leading
groups widiin the cities, alongside the urban nobility. In the maritime cities,
especially in Venice, the nobility itself participated in trade. But it is difficult to
get a picture of die social origins of merchants. Some of them were free, such
as the Cremonese milites active in the Po trade. But links with the bishops
appear repeatedly in die sources. Otto III and Conrad II gave the bishop of
Asti (at the moudi of the valley of Susa, one of the most important Alpine
crossings) in 992 and 1037 respectively a privilege granting freedom from tolls
to his merchants and to the citizens of his city. 27 Regardless of whether they
were free citizens or trading agents of the bishop, merchants profited more
than all other sections of die population from market, mint and toll privileges
and from the episcopal protection guaranteeing their trade-routes. Their room
for economic manoeuvre was defined by their link with their civitas and its
bishop. Archbishop Aribert could therefore justifiably be praised on his death
as mercatorum protector : 28 Nevertheless, the relationship was not always free from

25 Cf. Renouard (1969), p. 382. 26 D Lothario. 27 D O III 99; D C II 245.

28 Landulf Senior, Historia Mediolanensis n, 32.

Cambridge Histories Online © Cambridge University Press, 2008

Merchants, markets and towns

tensions, which show the importance of this group. In Cremona, disputes
between die merchants and the bishop are recorded as early as 924, when the
merchants sought to move the harbour to a different location to escape episco-
pal control. Tensions between citizens and bishops increased around 1000: in
983 there is a reference to conflicts between Milanese citizens and Archbishop
Landulf, and from 996 there were again disputes over the harbour and die
passage of ships in Cremona; these lasted a long time and broke into violent
conflict in 1005 and 1030. But the part played by merchants in these distur-
bances is not clear, and they really belong to die general wave of strivings for
autonomy which culminated in the valvassores' uprising in 1035. Nevertheless, it
is significant that evidence for the right of citizens to participate in the running
of the cities refers to economic affairs affecting merchants. Already in 948 King
Lotliar had granted the bishop of Mantua the mint, with the provision that the
conventus of the citizens of Mantua, Verona and Brescia was to determine the
fineness and weight of die coinage. This strong position of the citizens found
around the middle of die century was not seriously affected by the privileges
granted by the Ottomans to the bishops, and the economic well-being of die
cities was an important precondition for the formation of the communes in
the later eleventh century. By the beginning of the elevendi century at all
events, Italy ranked as the most advanced urban region of Europe. Her most
important cities were Milan and Venice, while the harbour cities of Pisa and
Genoa were rapidly gaining ground now that the Saracen danger in the
Tyrrhenian sea had been eliminated.

In the transalpine regions of the former Frankish empire, in what were
becoming France and Ottoman Germany, the development of towns took a
quite different padi. The wider mesh of the network of civitates (see above) left
room for further setdements resembling the episcopal sees in economic and
governmental function. But even those civitates going back to Roman times
operated under different preconditions from those of the episcopal cities of
Italy. The Germanic incursions at the end of the third century had led to the
fortification of the Gallic cities and so to a drastic reduction in die areas of
urban settlement. Only a few episcopal cities retained a substantial area: Lyons
(65 ha); Poitiers (47 ha); Rheims (60 ha); Sens (43 ha); Toulouse (90 ha).
Remarkably, these included some which lay near die limes and set up their
defences early: Cologne (96.8 ha); Mainz (98.5 ha); Metz (60 ha); Augsburg (61
ha, though here the fortifications had disappeared by the tendi century and had
no induence on the medieval development of the city) . Trier, the former impe-
rial residence, had the exceptional area of 285 ha, but only about 1 5% of this
was setded at die beginning of die tenth century. Most of the civitates had an
area between 6 and 1 5 ha: Auxerre (6); Limoges (7); Clermont (6); Le Mans
(7-8); Paris (1 5); Rouen (14). The areas in the towns widiin what later became

Cambridge Histories Online © Cambridge University Press, 2008

PETER JOHANEK

Germany were generally larger: Strasbourg (18.5); Worms (23); Regensburg
(24.5); Speyer (14). The smaller areas predominating in Gaul evidendy pro-
vided a model for the bishoprics newly founded in the Carolingian era in the
previously townless regions to the east of the Rhine, especially in Saxony:
Minden (4.24); Munster (7); Osnabriick (5.25); Paderborn (6.1).

It is clear that these civitates essentially had the functions of a mere citadel,
and the Old High German glossing of civitas and with pure underlines this

fortified character of urbanism. Market, trade, and to a large extent craftwork,
largely took place outside these fortifications. The settlements connected with
them were adjacent, but legally distinct, creating die characteristic picture of a
bi- or multipolarity in the early phase of town formation, which ended only
with the creation of a unified town law and wall-building enclosing the separ-
ate settlements in the course of the eleventh and twelfth centuries.

The process of town development was roughly similar in France,
Lotharingia and Germany, but the pace varied. The trade flows of the period at
first favoured Germany and Lotharingia: the links already mentioned across
the Alps with Mediterranean trade, which led into the Rhine valley or the
region of die Meuse and proceeded along these rivers to the coast; but also
the extraordinary growth of the slave trade within Europe. Regensburgwas the
crucial centre on the Danube route, Erfurt on the Thuringian, Magdeburg on
the route across the Elbe. It is no coincidence that it is in Regensburg that we
find around 1020 a civis and merchant of Slav origin settled there: Penno filius
Liubuste , 29

However, the most important impulse for development seems to have been
given by the intensification of trade across the Baltic, a flowering of the seeds
sown in the Carolingian era. Viking raids functioned here as a motor rather
than as a destructive force. The coastal region of northern Europe was bound
in this way into the network of long-distance trade; along the Dnieper and the
Volga a second trade-route was established with Byzantium and the Islamic
east. This Baltic trade also entered via the Rhine, Meuse and Scheldt estuaries
to end in Germany and Lotharingia and provided a significant economic
thrust. France, by contrast, remained cut off from Mediterranean trade by the
hindrance to trade via the mouth of the Rhone due to the Saracen threat; the
transit trade of Jewish and Verdunese merchants with Islamic Spain did not
compensate. Admittedly, Italian merchants are recorded around 1000 at the
Saint-Denis fairs, which go back to the Merovingian era, but the decisive rise of
the Lendit took place in the second half of the eleventh century.

One also has the impression that the French civitates only gradually recovered
from the depredations of the Viking raids, to which they had been exposed

29 Die Traditionen des Hochstifts Regensburg und des Klosters S Emmeram , no. 327, pp. 246—7.

Cambridge Histories Online © Cambridge University Press, 2008

Merchants, markets and towns

particularly strongly, even in the interior, during the period of the ‘great army’
from the 8 80s dirough to die foundation of Normandy in 91 1. In Paris, for
example, there is noticeable growth in the settlements outside the walled lie de
la Cite on bodi sides of the Seine only after the middle of the century; in
Rheims the churches of Saint-Denis and Saint-Nicaise still lay in ruins in the
mid-eleventh century. In Bordeaux the rebuilding also took place only at the
end of the tenth century, and not until the eleventh is a noteworthy tiow of
trade on die Garonne again recorded.

Nevertheless die London toll regulations of fEthelred the Unready of
984/ 5 mention long-distance trade widi the northern coast of France, espe-
cially with Rouen (wine and whale meat), but also with the mouth of the
Somme ( Ponteienses , die men of Ponthieu). 30 Yet in general it seems clear diat
the final phase of Viking raids retarded French development, whereas overall
and on balance they were a favourable impulse to north European trade, and
indeed played a decisive part in building up a trade network in the North and
Baltic seas.

Tenth-century France also lacked the driving force of powerful kingship.
Although the development of towns and markets in France, Lotharingia and
Germany was strongly influenced by regional political forces, the Ottoman
rulers played a decisive part. Their diplomata suggest that they had a trade
policy, one which was to intensify the impulses proceeding from die favourable
geo-economic conditions of the period. Their aim was to fill the area with
markets, places at which goods could be exchanged in ordered legal circum-
stances. The need to establish such places in particular regions is explicitly
stated. Kings themselves had such markets in their palaces and royal estates, in
the civitates and elsewhere. From the reign of Otto I the crown increasingly
granted the income from such markets in whole or in part, or the markets
themselves or at least the right to erect and run such markets, to odier lords.
Neverdieless, it continued to regard itself as a central regulator, for example in
the way in which it issued prohibitions against erecting markets in particular
areas, in order to protect the catchment areas of existing markets (as Otto III
did for Quedlinburg in 994), 31 but above all in the way in which it sought to
guarantee unitary principles of market law and custom. The charters granting
rights defined them by reference to those of die nearest economically signifi-
cant civitas (Cologne, Mainz, Magdeburg, Trier, Cambrai, Strasbourg, Speyer,
Worms, Constance, Augsburg and Regensburg) or odier royal market
(Dortmund, Goslar, Zurich). It is clear here diat we are dealing with royalhssr, as
when for example Otto I in his privilege for Bremen of 965 speaks of die law
of merchants in the remaining royal cities ( urbes ) and Henry II grants in 1 004 to

30 iViEthelred 2, 5— 6, in Die Geset^e dersingelsachsen, ed. Liebermann, 1, p. 232. 31 DO III 15 5.

Cambridge Histories Online © Cambridge University Press, 2008

PETER JOHANEK

the market Rincka in Breisgau the peace which is usual ‘in the greater places
and towns of our empire’. 32 Unified law held together a network of markets
owned by different lords, differentiated according to size and distance. The
granting of privileges to merchants themselves was a more immediate way of
encouraging trade, and this also occurred, though only a few traces of it have
survived. The merchants of Tiel, the successor to the Carolingian emporium of
Dorestadt, claimed at the beginning of the eleventh century to hold royal priv-
ileges, and Otto II had already granted tire merchants resident in Magdeburg
freedom from tolls throughout the kingdom with the exception of those at
Mainz, Cologne, Tiel and Bardowiekd 3 This clause, very much in the
Carolingian tradition, underlines once more the importance of the great
emporia on tire Rhine, the entry-points for the North Sea and Baltic trade. It
also sketches tire radius of action of a group of merchants in Magdeburg,
whose members are indeed traceable in Tiel.

What characterises diese merchants is their residentiality, their links with a
particular place, which is stressed occasionally in the diplomata, reflected in
phrases like Maguntinus institor or I erdunenses mercatores or deducible from their
sometimes considerable landed possessions, as when the Regensburg mer-
chant Wilhelm gives land in five different villages to die monastery of St
Emmeram in 983 . 34 This merchant residentiality also shaped die topography of
mercantile settlements and encouraged the formation of social groups with
permanent structures.

True, it is clear from the sources that merchants lived both inside and
outside the rivitas, as at Merseburg or Regensburg, but the setdements outside
the rivitas, known as suburbium, burgus, vicus or portus, took on a special impor-
tance. They were established, sometimes several of them, not only around rivi-
tates but also at palaces and royal estates, monasteries and aristocratic
fortifications. This consolidation of a vocational group will have encouraged
the formation of unions of a cooperative nature. The Magdeburg merchants -
occasionally named together widi the Jews - received dieir privilege as a corpo-
ration. 35 In Tiel, where the vicus ad portus of the merchants lay along two lordly
setdement-cores - the Walburgis monastery (an aristocratic foundation
granted by Otto I to the bishopric of Utrecht) and an important royal estate,
which was granted to St Mary in Aachen in 1000 — the outlines of a merchant
guild become visible, a ‘free association widi self-determined law for the pur-

32 D O I 307; D H II 78: ‘sicut in maioribus nostri regni locis et civitatibus’.

33 DOII112.

34 Liudprand, Antapodosis vi, 4 and 6, pp. 1 5 3—4, 1 5 5—6; Liber miraculorum S. Bertini abbatis, AASS
Septembrii n, cols. 595—604: ‘Viridunenses negotiatores’; D O II 293, cf. Traditionen des Hochstifts
Regensburg , no. 21 2, p. 192: ‘urbis Regie negotiator nomine Adalhart’.

35 DOIIii2;DOI 300.

Cambridge Histories Online © Cambridge University Press, 2008

Merchants, markets and towns

poses of mutual protection and support’, 36 which should be seen as a mile-
stone along the road leading to later inner-urban confraternities. Tiel’s espe-
cially vulnerable position at drat time may have encouraged the formation of a
guild, but similar associations may be assumed to have existed at other places.

Not all diose who traded would have fallen into the categories of mercatores ,
negotiatores, emptores and institores. The Raffelstetten toll ordinance distinguished
the Bavari (that is Bavarian landowners) trading in salt from the merchants and
Jews who to some extent were active in the same markets. 3 ' The distinction
between different groups of traders is difficult, but the inhabitants of via and
suburbia were probably characterised by their activity in long-distance trade.
Their social classification is equally difficult. The rich Regensburg merchant
Wilhelm (see above) had been ‘granted his freedom’ by the king. 38 There were
thus free men among the negotiatores , but it also means that others were active as
agents of the king and in his service, in bonds characteristic of what were later
to become ministeriales. Similar bonds are to be assumed for merchants in the
entourages of other lords, though their activities will have assured them a great
deal of flexibility in their legal status and way of life, something which enraged
monastic observers like Alpert of Metz, who describes the merchants of
Tiel. 39

The populations of these multiple settlements are frequently described in
terms which suggest that they were acting together, especially in conflicts with
the bishop of the civitas. It was the Metenses who blinded the bishop of Metz
installed by Henry I in 924. 40 But even when in 958 the citizens of Cambrai
sought to drive their bishop from the city, ‘united in one and tire same will and
having made a unanimous oath-taking’, 41 we are not yet dealing with an incipi-
ent citizens’ collective. We must reckon rather with different groups, legally dis-
tinct from one another, even within die civitas, in which often enough bishop
and count were in rivalry. In Cambrai the count held half of the town area and
of the dues; in several French civitates (e.g. Soissons and Amiens) comital castles
are recorded, and the Life of Bishop Burchard of Worms (1000—25) describes
impressively how Duke Otto, a son of Conrad die Red, possessed a fortifica-
tion within die civitas , which offered support to those persecuting the episcopal
familia. Bishop Burchard countered by fortifying the episcopal residence, and
so in time brought peace to a city in which widiin a single year thirty-five
members of the episcopal familia had been killed. But Burchard’s estate law,

36 D 0 1 124; D O III 347; cf. Oexle (1989), p. 1 84.

37 MGH Cap ., no. 25 3, n, pp. 249—52. 38 D O II 293.

39 Alpertus Mettensis, De diversitate temporum 11, 20— 1, pp. 78—82.

40 Adalbert, Reginonis Continuatio, s.a. 927, p. 158.

41 ‘cives una eademque voluntate collecti, factaque unanimiter conspiratione’: Gesta episcoporum
Catneracensium 1, 8 1 , p. 43 1 .

Cambridge Histories Online © Cambridge University Press, 2008

84 PETER JOHANEK

which set fixed legal norms within the civitas, applied only to his own familia, not
to other groups of persons.

In general the bishops were able to gain the upper hand within the area of
the walled civitas , not least with the help of the privileges granted them by the
Ottomans, and in the palaces, royal estates and aristocratic fortifications which
were also adjoined by via and suburbia the issue did not present itself. The
events just noted do make clear, however, that civitas and castle were centres of
lordship, though their significance cannot be confined to the merely military.
They were also far from serving exclusively as places of refuge for times of
war, even if Viking raids and Magyar razzias encouraged the building of fortifi-
cations. The Vita Burcbardi says that after peace had been established the cives
returned to live there. 42 That is understandable in the case of a civitas of the size
of Worms (see above), but even in very small settlements, such as that of the
castle of the counts of Flanders in Ghent (4 ha) of around 940 or 950, archae-
ology suggests that craftsmen were working there.

It must be stressed that in multiple settlements around civitates and castles,
lordship and fortification were closely linked. It was the legal form of the
market which proved attractive for tire exercise of lordship as well as holding
together the individual settlement cores of a civitas. Spiritual communities as
well as secular magnates set up markets, not least because they saw in them a
possibility of selling the agrarian produce of tire manorial economy. On occa-
sions this could cover quite a wide area. The monastery of Corvey on the
Weser grouped its scattered peripheral possessions by setting up markets
(Meppen in 946, and Horhusen (Niedermarsberg) by the beginning of the
eleventh century). 43 The most impressive example is that of Lorsch, which
intensified lordship in its neighbourhood by establishing a circle of markets
about 30 km away (Bensheim 956; Wiesloch 965; Zullestein 995; Weinheim
1000; Oppenheim 1008). These looked in part to the Rhine, in part to the
Odenwald, and show us that the region was receptive to commercial exchange.
Although tire bishopric of Worms, tire monastery’s great rival in developing
the Odenwald, had been able to concentrate large-scale trade on the market in
its civitas, it had only been able to penetrate the Odenwald itself with a single
foundation, Kailbach (1018), and here too only on the periphery. The record of
a settlement arranged by Henry II in 1023 shows that there had been a real
trade war, escalating at times into violence and even killing. 44

Secular magnates also made use of this combination of economy and lord-
ship, though records of their activities are much less well preserved. Count
Berthold, the ancestor of the Zahringer, set up a market in Villingen in 999,
and around the same time Otto Ill’s fidelis Aribo established markets in

42 Vita Burcbardi , c. 6, MGH SS iv, p. 835. 43 DD 0 1 77, 444. 44 D H II 501 .

Cambridge Histories Online © Cambridge University Press, 2008

Merchants, markets and towns

Donauworth, whose later history shows that they were also intended to have
functions of lordship. 45 The clearest case is the striving of the counts of
Flanders, not only in their building of castles in Bruges and Ghent provided
with portns and vims, but above all Count Arnulf ’s seizure in 9 3 9 of the fortified
oppidum Mentreuil at the Pas de Calais. What was at stake in this castle belong-
ing to Count Erluin was not merely its value as a fortification but also the ‘great
revenues’ which ‘were to be derived from die landings of ships’. 46

One may say in summary that the tenth century saw the opening of the
countryside for the exchange of goods at markets. This is true of Germany,
Lodiaringia and France equally, though in the west die process is not visible in
such detail, since the west Frankish king did not develop a market sovereignty
like that of the German ruler. The establishing of new markets seems to have
reached its height in the period around 1000. However, not all the markets set
up in the tenth and early eleventh centuries developed into towns. Many disap-
peared or acquired town law only very late. The market is therefore not die root
of die medieval town, but it prepared the ground for the urban economy and
can be described as the motor which kept die economic cooperation between
the separate setdement kernels going in this decisive phase of transalpine
urban development.

Overall we may assume very strong growth for the civitates and quasi-urban
setdements during the tenth century; occasionally this is visible in topographic
development. Thus in Regensburg — perhaps already under Duke Arnulf, cer-
tainly before 940 - an area to the west of the Roman legionary camp evidendy
setded by merchants was taken into its fortifications, effectively doubling the
surface area of the civitas to about 5 5 ha. In Worms a wall begun in the second
half of the tendi century was completed under Bishop Burchard, and the area
of setdement nearly trebled, from 23 ha to 65 ha. In Cologne die land won by
filling in die Roman harbour and setded by merchants was fortified around 940
or 950, which increased the civitas to 122 ha. Even on the smaller stage of the
royal estate at Dortmund, an important trading-centre nevertheless, we find an
increase in area from 2. 1 3 to 1 1 . 5 ha.

Growth of this kind certainly did not take place evenly everywhere, and we
must assume a certain hierarchy within the network of civitates, markets and
newly founded non-agrarian setdements near casdes and palaces. This is occa-
sionally mentioned explicitly, as in a letter of Abbot Odielbert of Saint-Bavo,
which describes the castrum Ghent as a caput regionisvAdch has precedence over
other civitates (here to be understood as a castle with a vims, not as an episcopal
city). To justify his view die abbot pointed to the church buildings and relics

45 D O III 3 1 1 ; D C II 144.

46 Richer, Historiae ir, 1 1 , p. 144: ‘eo quod ex navium advectationibus inde plures questus proveniant’.

Cambridge Histories Online © Cambridge University Press, 2008

PETER JOHANEK

there. 47 Elsewhere also we can observe how both bishops and secular mag-
nates underlined die importance of these places combining military installa-
tions, craft and mercantile settlements, and markets by endowing diem with a
special architecture and sacrality.

The building of walls was necessary for defence and had already been fos-
tered by the Viking threat of the ninth century. In the tenth century it grew and
became more effective, drawing in part on labour services from the agrarian
surroundings. But the building of new churches, especially new cathedrals
within the civitates, and their equipping with relics served display purposes and
encouraged streams of pilgrims, for whom, as a diploma of Otto Ill’s for the
monastery of Selz put it, a market was as necessary as for the monks and
the other people living there. 48 We can thus observe a lively building activity in
the tenth century, from Otto I, who endowed die cathedral he had built in
Magdeburg with a very rich set of relics, through bishops and abbots to nobles,
who added monastic or canonical foundations to their castles and also pro-
vided these with relics, like for example Manigold, the descendant of die fidelis
Aribo, who, in order to display more effectively the particle of the Holy Cross
which Romanos III had presented to him in Constantinople in 1029, comple-
mented his father’s market foundation in Donauworth with the foundation of
a spiritual community.

Large-scale buildings, the monasteries and collegiate churches founded
diere and their collections of relics increased the attractiveness of these places
both for secular vassals and for merchants, who found groups of wealthy con-
sumers to provide for. Besides the general economic conditions and the
impulses from lordship, the development of an impressive architecture and the
enhanced presence of die saints in their reliquaries belonged to the important
factors driving on the emergence of the medieval town in Germany and
France.

In the Mediterranean region and in the transalpine sections of the Frankish
empire the development of urbanism was shaped by lines of continuity leading
back into antiquity, even if these were absent in the easternmost part of the
empire. Northern and eastern Europe could not build on such traditions, and
even in Britain, where England and Wales had a Roman past, these traditions
were not effective to the same extent as on the continent. For the whole of this
region, with die exception of England, it must also be acknowledged that our
knowledge of urban development owes much more to archaeological research
than to written sources. Even in England, the archaeological investigation of
towns is further advanced than it is on die continent.

47 Elenchusfontium historiae urbanaeii , 2, no. 8, p. 295.

48 D O III 130: ‘et mercatus necessaria sunt multitudini populorum undique illuc confluentium, simul

etiam monachis et populis ibi commanentibus et habitantibus’.

Cambridge Histories Online © Cambridge University Press, 2008

Merchants, markets and towns

From the early middle ages onwards England played an important part in
North Sea trade, which was of such importance for the economic develop*
ment of Europe precisely in the tenth century. The Viking raids and
Scandinavian setdement brought it, together with the other parts of Britain
and Ireland, into still closer contact with Scandinavia, to the point where it
became part of a Scandinavian empire under Cnut and his sons.

In England, the monk TElfric referred to merchants in positive terms,
describing and defining their activities in his Colloquy , shordy before Alpert of
Metz made his harsh judgement on die merchants of Tiel. 49 It is from England
also that we have the earliest evidence for medieval European merchants’ own
thinking, and this in turn shows just how far Scandinavia and the Baltic region
lay widiin die ambit of Anglo-Saxon kings. At the court of King Alfred die
Great (871-99) the Norwegian Ohtere described his journeys to the Lapps and
to the coasts of Norway and Denmark along as far as Haidiabu; the Anglo-
Saxon Wulfstan related his knowledge of the Baltic from Haithabu to Truso on
the Vistula estuary and beyond into the lands of the Estonians. King Alfred
included diese reports in the Old English translation of the World Chronicle of
Orosius. 50

Alfred’s government marked an important turn in the development of
Anglo-Saxon urbanism. Until dien there had been in essence three kinds of
quasi-urban setdement in Anglo-Saxon England. The first consisted of centres
of royal power located within die walls of Roman cities - London, York,
Canterbury and Winchester - which were also bishoprics. It should be noted,
however, that die density of setdement within the Roman fortifications was
very low. The next type was that of unfortified trading emporia on the coast,
with names frequendy ending in -me: Hamwih (Southampton), Fordwich,
Sarre, Dover, Sandwich, Ipswich. The ending -wic also appears in Eoforwic
(later Scandinavian Jorwic = York) and Lundenwic (= London). Archaeology
has revealed that west of Roman London there was indeed an emporium of
this kind, widi an area of at least 24 ha, perhaps 80 ha, lying between what are
now Fleet Street and Whitehall, described by Bede as a significant centre of
long-distance trade. In York also crafts and trade seem to have been practised
mainly outside the area of the Roman legionary camp even before the
Scandinavian conquest of 862. It would seem that some of these trading
emporia formed a functional unity with nearby royal centres: Hamwih with
Winchester; Ipswich with the region around Woodbridge, Rendlesham and
Sutton Hoo. A third group was made up of newly established fortified setde-
ments in the interior, such as the Five Boroughs of die Danelaw (Stamford,
Nottingham, Derby, Lincoln and Leicester) or else Hereford, about which litde

v> /Elfric, Colloquy, pp. 3 3 — 4. 50 Tlse Old English Orosius, pp. 13— 18.

Cambridge Histories Online © Cambridge University Press, 2008

PETER JOHANEK

is known archaeologically. Overall, however, there were probably not more
titan about fifteen settlements with urban characteristics in Alfred’s time.

English scholarship has established a bundle of criteria to determine what
marks a town in Anglo-Saxon England: market, mint, fortifications, tenements
and open fields (Stenton), special jurisdiction (Lyon). 51 These criteria apply to
the majority of a group of settlements which become visible under Alfred and
his son Edward die Elder and are linked with a defensive plan directed against
die Danes. The Burghal Hidage, a list dating from around 914-19, names 30
burbs in Wessex and three in Mercia: 52 fortified locations, which were to be kept
in repair by the surrounding population and could be manned in times of
danger. We are dealing here in part with die use or reuse of Roman or even Iron
Age fortifications, but mostly with new settlements. This defensive system,
which could also play a part in attack, was the basis for resistance to the Great
Army of the Vikings which turned against England from 892 onwards; its
development accompanied die recuperation of the Danish north.

To make these settlements capable of surviving and functioning they were
mosdy equipped with a mint and a market, with the latter appearing in royal
legislation as port. From the time of Edward the Elder sales were restricted to
the port and dierefore to the burb, where they were to be supervised by a royal
official, the port reeve, and made before market witnesses. When Alfred occu-
pied London in 886 he evidendy caused the vicus on the Strand to be incorpo-
rated into the walled region; in similar fashion die area of the former Roman
town in Winchester was filled with setdement. Economic function and fortifi-
cation came into line with each other. In the interior bodi new settlements and
ancient urban locations were given a regular street network, so diat we can
speak, with Biddle, of ‘planned towns’. 53

The kingdom of Wessex thus covered the country with a network of for-
tified markets, which in their function were comparable with the markets of
Ottoman Germany but were all controlled by the king. Their legal and topo-
graphical form made diem die basis for the medieval English borough, even if
- as on the continent - not all the settlements of the tenth century dourished,
or are still found as boroughs in Domesday Book. Places like Halwell and
Chisbury remained mere hillforts; some, like Gothaburh, cannot even be iden-
tified with certainty. On the other hand we can already find in the tenth century
a fundamental difference between town and rural settlement, as soon as
‘greater population numbers, walls, market, mint, income of the population
derived partly from trade and craft, market witnesses, royal officials and courts’
come togedier. 54 In the laws, provisions regulating urban conditions become

51 Haslam (1994), p. xv.

52 Hill (1969). 53 Biddle and Hill (1971).

54 Die Geset^e dersi ngelsachsen, ed. Liebermann, n, p. 660, 1 h.

Cambridge Histories Online © Cambridge University Press, 2008

Merchants, markets and towns 8 9

more and more frequent. In Anglo-Saxon England there was also a hierarchy
of towns and markets, and for diat reason concentrations of commercial activ-
ity. The activities of royal mints provide a barometer. Their numbers had risen
sharply towards the end of the tenth century, from twenty-seven and later forty
under Edgar to seventy-five under fEthelred the Unready, while Domesday
Book notes eighty. The laws of King dEthelstan prescribe a mintmaster for
every hurt), but there were exceptions: London 8, Canterbury 7, Winchester 6,
Rochester 3, Lewes, Southampton, Wareham, Exeter and Shaftesbury 2.
dEthelred tried to reduce the increased numbers, but allowed each summits portus
(principal town) three. 55 And indeed we find in the various regions places with
an above-average mint output (London, Winchester, York, Lincoln,
Canterbury, Exeter, Chester and Norwich), accounting together for more than
half of total output. It was the south and east which dominated here; in the
north only York, with 9% of total production, stood out.

The evidence of written sources, archaeology and numismatics suggests a
lively urban life and internal trade, with commercialised forms of goods
exchange between town and countryside. Naturally, England was by no means
isolated from continental long-distance trade. The London trade regulations of
dEthelred reveal the close links with die neighbouring coasts across the
channel. Besides the merchants from France and Flanders, mentioned above,
we find die ‘men of the emperor’, German traders. 56 Among these those of
Huy, Liege and Nivelles are given particular attention; presumably they traded
in bronzeware. fElfric’s merchant, mentioned above, deals in goods which
point still furdier afield: purple and silk, valuable stones and gold, various
clothes and spices, wine and oil, ivory and golden bronze ( auricalcum ), iron ore
and tin, sulphur, and glass. 57 These names recall the routes over the Alps which
brought the Anglo-Saxons to Pavia, where they exchanged their goods for
purple dye, silk, spices and other things.

The travellers’ accounts of die merchants Ohtere and Wulfstan included by
King Alfred in his translation of Orosius point equally definitely to die Baltic
and die activities of the Scandinavian peoples, however. They show that die
British Isles and the southern Nordi Sea coast of die continent formed a
system of trading emporia: Dublin, other Irish sites and York in the west; in the
Baltic region Kaupang in the fjord of Oslo, Haithabu in die bay of Schleswig,
Birka in the Malar region of Sweden, Paviken and other locations on Godand,
and the sites of the soudiern coast of the Baltic - Ralswiek on Riigen, Wolin
(Jumne) and Menzlin around the mouth of the Oder, Kolberg (Kolobrzeg) on
the Pomeranian coast, Truso in the delta of the Vistula, Grobin in Kurland and

55 Die Geset^e dersingelsachsen , ed. Liebermann, i, n /Ethels tan, 14, 2, pp. 1 5 8—9; iv/Ethelred 9, p. 236.

56 Die Geset^e der Angelsachsen, ed. Liebermann, 1, iv /i '.thclred, 2, 8, p. 234.

57 Elenclms fontium historiae urbanae , no. 8 , p. 295.

Cambridge Histories Online © Cambridge University Press, 2008

9°

PETER JOHANEK

Daugmale on the lower reaches of the Dwina. These sites date back to the
eighth and ninth centuries, though they had undergone extensive and very var-
iable change in the course of the tenth century. But all of them had been given
new impulses by the Scandinavian expansion from the ninth century onwards,
seen most obviously in the razzias and trading missions of the Vikings.

This line of trading sites from west to east pointed towards those sites which
organised trade with Islamic central Asia and Byzantium via the Russian rivers.
In the tenth century the most important centres at first were Staraja Ladoga on
the Volkhov and Gnezdowo (the precursor of Smolensk) on the Dnieper. The
first could be reached by ship across the Gulf of Finland and the Neva, and
opened up routes to both Dnieper and Volga. The latter could be reached more
directly from the Baltic via the Dwina. Gorodisce on the Volkhov, in die course
of the tenth century to be gradually replaced by Novgorod, 2 km further south,
and Kiev on die Dnieper, the heart of the Rus' empire, should also be men-
tioned. Scandinavians were present in all these sites, and indeed played a crucial
role. Haithabu, Kaupang and Birka were controlled by Scandinavian kings;
Dublin originated in 917 as a Viking foundation, and Scandinavians were of
decisive importance in the development of the Russian sites as well, even if
details are disputed. The Irish Sea, tire North Sea and the Baltic together could
count as a Scandinavian sea. The Icelandic sagas reflect this; Egil’s Saga, for
example, calls die journey to Dublin ‘the most popular route’, 58 and Egil and
his companions are shown visiting Norway as well as Wolin and die coast of
Kurland. These trading emporia, especially in the Baltic and in Russia, were
generally polyethnic formations, rather like the one at Birka described by
Adam of Bremen, though in his own time this had long ceased to be impor-
tant: ‘all the ships of the Danes, the Norwegians, the Slavs, the Sembs, and
other Baltic Sea tribes are accustomed to assemble there regularly to pursue
their necessary affairs’. 59

Compared with the older Viking era in the ninth century there was consider-
able growth in these centres in the tenth century, coupled with new founda-
tions and shifts in site: Dublin and Novgorod, as we have seen, and one might
also name Sigtuna, which took on Birka’s role in the Malar region from about
980 onwards. Only one of diree settlements continued in existence at
Haidiabu, but this grew in the course of the century to a size of 24 ha. It was
precisely the most important sites which displayed such growth: Wolin grew to
20 ha, Staraja Ladoga grew from 4-5 ha to 10 and Gnezdowo from 4 to 1 5 ha.
Mostof the other sites lay between these last two (Birka 13, Dublin i2,Menzlin
10, Ribe 10); some, like the oldest settlement at Danzig (1) or Daugmale (2)
were much smaller. Wolin and Haithabu thus headed the league table.

58 Egils Saga Skallagrimssonar, c. 32, p. 100 (English trans. p. 82).

59 Adam of Bremen, Gesta 1, 60, p. 5 8.

Cambridge Histories Online © Cambridge University Press, 2008

Merchants, markets and towns 9 1

The growth of these sites points to an intensification of exchange, and the
composition of the coin hoards, which in the Baltic regions show a high pro-
portion of Arab silver coins up to about 970, suggests that till then the west
had a trade surplus, even if some of tire dirhams which came westwards were
derived from tribute -payments to the Varangians. Western exports certainly
included woollen cloth, mentioned by Adam of Bremen and confirmed by
archaeology in Birka and Wolin. Wine evidently also reached at least as far as
the Baltic in considerable quantities. The most important role was probably
played by tire slave trade, however, whereby the product of Viking razzias in
western Europe (Ireland in particular) was marketed in Scandinavia and the
Muslim east. Haithabu and Brenno at the mouth of the Gotaalv, as well as the
Volga, are noted as points on this trade-route. The treaties between the princes
of Kiev and Byzantium in the tenth century also mention slave-trading; die
Russian regions were also a source of slaves. The high proportion of total trade
made up by slaves is the most coherent explanation of the import of Arab
silver and other wares (silk, for example, has been found in Birka, Wolin and
Dublin). In the last quarter of die tenth century the structure of trade relation-
ships changed. From now on western silver tiowed towards the Baltic and
Russia (see above, p. 66). This means diat Russian exports of raw materials
must have increased, most likely wax and furs, sought after by the west ‘as
much as eternal salvation’. 60 By contrast, the export of Christian slaves from
the west to the Islamic east will have declined and dien stopped; the reasons lay
presumably in the monetary difficulties of the central Islamic realms, but also
in the gradual Christianisation of the Scandinavian kingdoms from 965 and the
increasing prosperity of north-west Europe.

Besides the principal items of trade - slaves, wax, furs and luxury goods - we
find a wealth of other raw materials and craft goods, which were traded in large
quantities over medium and long distances and marketed in the sites just men-
tioned: Rhineland glass and pottery, Scandinavian vessels of soapstone and
metalware. Trade intensified in the Baltic region as well, and there was a lively
exchange with the newly evolving market systems of the central European
regions, especially in the Rhineland and in Saxony. That explains the special
role played by Haithabu, the link to Scandinavian trade, but also by Wolin,
which was of similar importance for its links down the Oder with the emerging
lordships of central Sclavinia.

This intensification of trade in luxury consumer goods is certainly also the
explanation for the location of craftsmen in the trading-centres of the
North Sea and Baltic regions. Metal- and leather- working in particular can be
confirmed archaeologically at various sites (Dublin, Haithabu, Wolin, Birka)

60 Adam of Bremen, Gesta iv, 1 8, pp. 244—5 •

Cambridge Histories Online © Cambridge University Press, 2008

9 2

PETER JOHANEK

and grew in importance in the eleventh century. The location of crafts contrib-
uted to the growth and die thickening of settlement at north European trading
centres. In many cases these were wholly or partially fortified in the course of
die tenth century, and there are suggestions that there was a layout of fixed
plots. Nevertheless, these quasi-urban setdements were of very varied stability.
Kaupang in Norway was abandoned at the beginning of the tenth century;
Paviken on Godand and Menzlin vanished around i ooo, and Birka was gradu-
ally displaced by the royal centre of Sigtuna from about 970 onwards, while
Haidiabu was replaced in die elevendi century by Schleswig - the earliest
cadiedral may already have been built there in the time of Cnut. Dublin, the
seat of Irish Viking lordship, the episcopal seat of Ribe, and the Russian
princely towns continued in existence.

It is evident that lordship helped to stabilise economic centres as well as
drawing economic functions to it. This is most evident in the inland regions of
the west Slav peoples, where fortified towns like Gniezno, Cracow and Opole
in Poland, Teterow, Brandenburg and Starigard (Oldenburg in Holstein) in the
region between Elbe and Oder, or Kourim, Liubice and above all Prague in
Bohemia dominated the picture. These were multiple settlements including a
lordly fortification, a suburbium , and craft working, for which there is archaeo-
logical evidence. They also evidently played a role in trade. Prague in particular,
which as a centre of rulership came to surpass all other Bohemian fortifica-
tions in the last third of the ninth century, developed in the course of die tendi,
with its extended surburbium , the Mala strana on the Vltava, to an inland
trading-centre, in which, in die words of Ibrahim ibn Ya'qub, ‘the Rus' and the
Slavs from the city of Cracow’ and ‘Muslims and Jews from the lands of the
Turks’ came togedier in a polyethnic market similar to those of the maritime
trading-centres of the Baltic region. 61 In the Baltic the foundations for the
later-medieval trading history of northern Europe, shaped by the Hanse, were
laid in the nindi and tenth centuries. The emporia of die northern Baltic on
which diis development was based had an urban functionality, or at least ful-
filled in great part the roles characteristic of towns in later centuries. But in
diemselves they were mostly not the starting point of the urban development
of the high middle ages, tending rather to disappear again. Towns came into
being for the most part in places where markets and trade were linked with
centres of secular or ecclesiastical power, as in Dublin, Ribe, Sigtuna,
Novgorod and Kiev. That is equally true of inland Sclavinia, where although
emporia like those on the Baltic were unknown, casual markets are mentioned.
But no town emerged from the unlocalisable ‘market of the Moravians’ men-
tioned in the Raffelstetten toll-ordinance; 62 it was Prague, a centre of lordship,
which was to become a town.

61 Arabische Berichte, p. 12. 62 MGH Cap., no. 253, ii.249— 52.

Cambridge Histories Online © Cambridge University Press, 2008

Merchants, markets and towns

Thus it was that the north-eastern region of Europe, which contributed so
significantly to European economic development, had litde effect on die devel-
opment of the town in the high middle ages. It is also unclear how far tenth-
century conditions in this region contributed to the formation of that type of
long-distance trader and merchant which was to have such a strong influence
on die institutional development of towns in the eleventh and twelfth century.
Trade in the Viking era was carried out by merchants who were also often
active as raiders or as warlike conquerors demanding tribute. The written
sources give the impression that trade was in many cases only a part of their
economic activity (as for instance with the landowner Ohtere) or was practised
only for a part of their life-cycle. These traders operated in communities, as
runic inscriptions occasionally reveal. But we are evidently dealing with short-
lived and casual communities with no fixed location, not with long-term
unions bound by oaths as with the merchant guilds of the European continent
(see above). Such corporations are evidenced in Scandinavia, as in Tiel, only in
the runic inscriptions of the eleventh century. By contrast with western and
southern Europe we know little about the shaping and maintenance of die
market peace, or the self-organisation of the merchants, or the form and
extent of the influence exerted by princely power. We must therefore conclude
that essential features of the medieval town - both its social and juridic make-
up and its topography and visual image - were formed in the core of
Carolingian Europe, in the civitates of northern Italy, and of the west and east
Frankish kingdoms. In these civitates and settlements of similar structure the
importance of ministerial dependants of lords and of prosperous and increas-
ingly professional long-distance traders, with a tendency to form guilds and
settle permanently, grew in the course of the tenth century. Governmental
peace ordinances to regulate the market were conceived here. All this was an
anticipation of tire later distinctions between the legal and social spheres of
urban and of rural life, so that already around 1000 Notker the German could
contrast purclich Andgeburlich, ‘townly’ and ‘farmerly ’. 63 This distinction took on
its final form once institutional structures had been developed for tire social
formation of the town between the eleventh and the thirteenth century.

But it was above all tire growth visible everywhere and its associated building
activity which shaped the characteristic picture of the medieval European
town. It was the great stone buildings of the church and of rulers which were
decisive here, and these were being imitated in the Slavonic east even in our
period: Ibrahim ibn Ya'qub stresses that the city of Prague was built of stone
and mortar . 64 The equipping of civitates with a ring-wall and a multiplicity of
churches, often located according to a preconceived plan, was the manifesta-
tion of an urban ideal which lords gave architectural form. It modelled itself

63 Notker the German, Werke i, p. 1 1 1 . 64 Arabische Berichte^. 12 .

Cambridge Histories Online © Cambridge University Press, 2008

PETER JOHANEK

on the ‘Holy City’, as for example when Meinwerk of Paderborn is said to have
built his episcopal city ‘in die form of a cross ’. 65 The rich stores of relics and
the frequendy attempted or at least invoked imitation of die example of Rome,
the city of antiquity, are important elements of this tenth-century urban idea.
The picture of walls, churches and towers as a city’s ornatus belongs to the
inheritance bequeathed by the tenth century to the cities of the European
middle ages, who have preserved it as an abbreviated symbol of urbanity in
depictions on their seals.

65 ‘in modum crucis’: 1 Vita Meinverci episcopi Patherbrunnensis , c. 2 1 8 , p. 1 3 1 .

Cambridge Histories Online © Cambridge University Press, 2008

CHAPTER 4

RULERS AND GOVERNMENT

Janet L. Nelson

IMAGES OF MONARCHY IN THE TENTH CENTURY

Tenth-century churchmen emphasised the kingship of Christ, and made king-
ship Christ-centred. They called on the earthly king to be Christ’s special imita-
tor. Like Christ, the king must willingly undergo travails: like Christ’s, a king’s
service, even his humiliation, brought glory to him and well-being to his
people. The theologians were also preoccupied with Antichrist. They pon-
dered the end of time, scanning their natural environment for supernatural
signs and portents. All of them believed that they lived during the Last of the
Four Empires predicted by Daniel in the Old Testament. The Last World
Emperor would, according to prophecy, establish a reign of peace, vanquish-
ing the enemies of Christ. Then would follow die Last Days: die brief rule of
Antichrist, and the Second Coming of Christ himself. These learned men were
ecclesiastics, many of them monks. But they were in close contact with the
secular world, and among the leading counsellors of kings and queens. When
the learned produced political thought, monarchy dominated their specula-
tions. Kings were frequently their addressees.

Monarchy could take the form of empire. Only in Italy perhaps, among
western lands, was there still a sense of Constantinople as die imperial centre
of the ‘Roman’ world. Elsewhere, imperial rule tended to be non-Roman, and
defined in terms of rulership over a number of realms. While the Carolingian
model inspired the Ottoman Reich, it had become clear by the close of the
tendi century diat die kings of the west Frankish realm recognised no imperial
overlord. In Poland and Hungary, Otto III, strongly infiuenced by such learned
tutors as Gerbert of Rlieims, may have seen himself as summoning a new
world of kings into existence to redress the balance of the old; but his attempt
at a Roman ‘renovation’ was bound up with ecclesiastical reform and had
eschatological dimensions. Kingship too could be viewed in apocalyptic per-
spective. Adso of Montier-en-Der, writing c. 9 50, believed that the kings of the
(west) Franks protected the world from Antichrist: ‘as long as their rule

Cambridge Histories Online © Cambridge University Press, 2008

JANET L. NELSON

endures, die dignity of die Roman realm will not wholly perish’. 1 Tenth -
century west Frankish kings could also be credited with a realm that was ‘impe-
rial’ because it consisted of several entities called regna (regions dignified as
such whedier or not diey had once been independent kingdoms). 2 England,
another imperial realm in that sense, was actually created in the tenth century.
In the 990s, when that new realm was subjected to renewed Scandinavian
onslaughts, the theologian dElfric, who believed himself to be living at the end
of time, celebrated Edgar’s far-fiung overlordship and preached die virtue of
obedience to divinely instituted monarchy. 3 His message was spread widely:
fElfric’s medium was English, and his audience included the local nobility as
well as monks.

Tenth-century historians - Widukind, Liudprand, Flodoard, Richer - pro-
duced powerful images of royalty. For these writers, all of them monks or
clerics, the deeds of kings continued to be the stuff of history, hence of moral
lessons. Classical models, especially Sallust, hovered behind these texts. Yet
Widukind seemed also to reflect his contemporaries’ confidence in the special
qualities of kings: their capacity to bring victory and well-being. Some twentieth-
century German commentators have heard echoes here of what they have
labelled Germanic notions of Heil\ Widukind drew, more certainly, on die Old
Testament. The lives of royal saints were another lively genre: in Ottoman
Saxony Queen Matilda, widow of Henry I, was venerated in a court-linked
monastic cult not long after her deadi, while in England die west Frankish visitor
Abbo of Fleury counterposed to the martyred King Edmund of the east Angles
(died 871) the martyr’s Viking persecutor as archetypical bad ruler. 4 Saint-kings,
and -queens, were depicted as holy not ex officio , but through special personal
qualities. While such ancestors shed charisma on descendants, none of this
should be seen as ecclesiastical flirtation with pre-Christian ideas of sacral king-
ship. Only some fairly heavy interpretation of Flodoard ’s and Richer’s accounts
of the deaths of successive tenth-century west Frankish kings has allowed
modern scholars to hypothesise popular belief in the last Carolingians’ loss of
royal thaumaturgic powers. 5 What the. Histories And the Lives alike conveyis a pro-
found confidence in Christian rulership, and the capacity of churchmen and -
women to construct potent images thereof.

Arbitration and protection were recurrent needs; and, even in regions where
kings were weak, or seldom if ever came, it was remembered - in Christian

1 Adso, Epistula ad Gerbergam reginatn , p. 26.

2 Richer, Historiae iv, 1 2, p. 162; cf. Hugh, charters 3,10, ( RHF 1 o, pp. 5 50, 5 60).

3 JE lfric, Lives of the Saints , pp. 468—70; partial trans. EHD, pp. 927—8.

4 Abbo of Fleury, l/ita Sancti Eadmundi , cc. 7—1 o, cols. 5 1 1— 1 5 .

5 Flodoard, Linn ales, s.a. 954, p. 138, Richer, Historiae hi, 109, p. 127, as interpreted by Poly and
Bournazel (1991), pp. 499—500.

Cambridge Histories Online © Cambridge University Press, 2008

Rulers and government

lands, the Bible and the liturgy offered constant reminders - that the provision
of justice and peace had been and still were tire function of kings. Especially in
the western parts of what had been the Carolingian empire, clerical and lay
ideals converged on the figure of Charlemagne, dispenser of justice and fol-
lower of wise counsel as well as mighty warlord. Several generations before the
earliest extant manuscript of the Song of Roland was written, Adhemar of
Chabannes wrote of the just rule of Charlemagne and ‘knew’ that he had
extended his realm as far as Cordoba. 6 Social memory, transmitted through
vernacular songs, perpetuated for the denizens of secular courts a vision of
monarchy capable of bearing apocalyptic expectations. Early in the tenth
century, King Louis of Provence, himself (like many leading figures) of
Carolingian descent on his mother’s side, named his son Charles-Constantine.
At the century’s close, in the year 1000, Otto III visited Charlemagne’s tomb at
Aachen. 7

Though theologians tended to measure by millennia, historians were
absorbed by the here-and-now. Flodoard of Rheims, who wrote of his own
church’s past in the expansive genre of History, used Annals for recent political
events, crowding each year with details of royal activity. Yet far beyond the
scholar’s study, and beyond Rheims, the existence of kings, and the legitimising
force of drat existence, were known and felt. Monks and clerics in Burgundy,
the Limousin, die Midi, used royal reign-years to date die documents of pow-
erful laymen. 8 The cults of saints particularly associated with the monarchy
were widespread. A tale recorded in the Life of Odo of Cluny shows St Martin
concerning himself direcdy with west Frankish royalty: a hermit somewhere in
the south of France saw him one day (it should have been 19 June 936) in a
vision - briefly, for the saint explained that it was the day of the king’s (Louis
IV’s) consecration and he had to be there at Rheims. 9 This may not be history,
but it resembles the work of the Rheims historians in keeping a whole realm in
view. The same is truer still of historiography in the east Frankish realm.
Widukind, Hrotsvitha, and the authors of the Quedlinburg Annals, based
though they all were in Saxony, narrated the deeds of kings and magnates
throughout the kingdom, with the itinerant court providing a strong central
focus.

Elsewhere, it was not historiography which promoted any realm-wide
vision. This is not surprising in the new and still essentially non-literate king-
doms of the tenth century: Denmark, Norway, Poland, Hungary. In all of
these, we shall see (below, pp. 107-1 2) the deployment of oral, visual and cultic
media. Kings found similar ways of fostering die image, and the social reality,
of a united realm in England too, and perhaps more surprisingly, an earlier

6 Adhemar, Chronicon n, 1, p. 68. 7 Thietmar, C/jromconiv, 47, pp. 184/6; Gorich (199S).

8 Kienast (1969). 9 John of Salerno, Vila Saudi Odortis, c. 27, col. 5 5.

Cambridge Histories Online © Cambridge University Press, 2008

9 «

JANET L. NELSON

royal interest in the maintenance and diffusion of a chronicle record was
allowed to lapse. In the Italian kingdom, there was, still, a state of sorts, but
there had not grown up, since the Frankish conquest, any tradition of royal
annals. Literacy, here relatively abundant, remained closely linked with law and
government, which turned out to be operable without kings and at local level.
Here, in the absence of any court-centred, realm-focused historiographical
tradition, laymen wielded power without looking to kings to legitimate it.
Where men no longer drought positively about kings, it was difficult to imagine
a kingdom. In this respect, Italy in the tenth century was different from odier
post-Carolingian lands. And even in Italy, but still more clearly elsewhere in
Frankish Europe, an ideal- type of Carolingian government was transmitted to
die learned through the written residue of capitularies, conciliar decrees and
documents. Manuscripts of capitularies continued to be copied in episcopal
scriptoria during die tendi century. In the 990s, Abbo of Fleury commended,
and cited at length, to Kings Hugh and Robert the conciliar canons with which
dieir predecessors Charlemagne and Louis the Pious had promoted the well-
being of both ‘state’ ( respublica ) and church. 10 Among the books consulted by
Bishop Fulbert of Chartres, when c. 1020 he wrote to Duke William of
Aquitaine on fidelity, was almost certainly a capitulary collection. 11
Sacramentaries and pontificals, in all the realms of the former empire including
Italy, continued to include prayers for kings. Law and liturgy, as well as songs
and stories, were forms of social memory in which Carolingian traditions sur-
vived.

In die course of the tenth century, the patrilinear dynastic link with the
Carolingians was broken in both west and east Frankish kingdoms: Carolingian
traditions, attached to kingship itself, could be taken up by Capetians as well as
Ottomans, and also were readily exported outside the old Frankish lands, most
notably to England where Edgar’s regime was a passable imitation of
Carolingian models. Contemporary writers who imagined, and tried to influ-
ence, the workings of kingship did not neglect royal marriages. As the parvenu
Henry I sought to extend his power into the Lotharingian heart of the
Frankish world and to achieve wide acknowledgement of his legitimacy, he
married his daughter to the leading Lotharingian magnate Gislebert and his
son Otto married die English princess Edith. Otto looked further afield to find
a bride for his son: when Otto II married the Byzantine princess Theophanu,
die Frankish world opened out to realms beyond it, prefiguring a wider
Europe. Foreign princesses brought the prestige of other royal lineages. Those
brides had the advantage of relative detachment from demanding aristocratic
families within the realm. Alternatively, kings would marry into just such fami-

10 Abbo of Fleury, Ubercanonum, PL 139, col 477. 11 Fulbert, ep. 5 1 , pp. 90—2.

Cambridge Histories Online © Cambridge University Press, 2008

Rulers and government

lies to gain die countervailing advantage of regional alliance and support:
Matilda, second wife of Henry die Fowler and mother of Otto I, was chosen
for her powerful Saxon connections; the west Frankish queen Adelaide, wife
and modier of die first two Capetians, was herself die daughter of the duke of
Aquitaine; while the English king Edgar married successively the daughters of
leading nobles in the west and south-east of his realm. The enhanced political
role of the queen - as Widukind said of Matilda, ‘she would sit while around
her the people stood’ 12 - was a striking feature of the tenth century. These
women were called dominaeP in translation, the pallid modern English ‘ladies’
hardly conveys something akin to lordliness in their authority. The Virgin could
be imagined as domina endironed, and even crowned. 14

TENTH-CENTURY KINGDOMS

In 888, the ‘old’ Carolingian realms had (as Regino put it) created kings out of
their own guts. 15 The pattern of the mid-ninth century was resumed: a three-
way division, with east and west Frankish kings vying for an increasingly frag-
mented Middle Kingdom. The later existence of die separate states of France
and Germany has tended to evoke assumptions of historical inevitability on
the part of French and German historians. Yet diere were forces at work in the
successor-kingdoms of Charlemagne’s empire which could have led to other
outcomes. On die one hand, Frankish unity remained a reality until late in the
tendi century. Eastern and western kings continued to vie for Lodiaringia;
some members of the east and west Frankish elites, and hence the kingdoms
they sustained, remained bound together, frontierless. The marriage of Otto
I’s sister to the west Frankish king Louis IV was at once symptom and agency
of a close entente drat came close to Ottoman hegemony in mid-century. On
the other hand, the regionalisation of the elite had gone a long way, and the
leading families of the regna most distant from the Lotharingian core - the
Billungs in east Saxony, for instance, or the Poitevin dukes of Aquitaine -
lacked any perceptible pan-Frankish perspective. Some of the old resources
were no longer available to tenth-century Frankish kings in east or west: lack of
ancient royal lineage was a drawback for Ottomans and Capetians alike, and
there was seldom much to offer in the way of plunder, tribute or territorial
expansion. Where earlier Carolingian rulers had sometimes been able to make
and, more rarely, break powerful men, by conferring, withholding and redis-
tributing comital office, this was manifestly more difficult in the tenth century
as countships and the lands that went with them were inherited over several

12 Widukind, Res gestae Saxonicaem, 74, p. 125. 13 Gerlx-.rt, epp. 62,66, pp. 6 1,64.

14 von Euw (1991), pp. 122— 4; cf. Deshman (1988). 15 Regno, Chronicon, s.a. 888, p. 1 29.

Cambridge Histories Online © Cambridge University Press, 2008

IOO

JANET L. NELSON

generations. Aristocrats’ own regional power always tended to become
entrenched, with house-monasteries and the privatised cults of saints repro-
ducing in their patrimonies the local domination effected by each royal
dynasty’s religious patronage in its heartlands. In east and to a lesser extent west
Francia, the identities of regions still dignified as regna within the realm
remained clearly etched. In both realms, the period c. 920 was especially critical:
at that point, it was no foregone conclusion that either would survive as a polit-
ical unit instead of falling apart into constituent regna. By c. 1000, however, the
survival of both was no longer in doubt.

This change — one of tire most striking features of the tenth century — is not
easy to explain. It can hardly be attributed solely to the policies and doings of
tire rulers themselves, given the relative lack of royal resources already noted
especially in west Francia. Rather, the focus needs to be widened socially to
include the aristocracy, for it was their perceptions of interest and of what con-
stituted legitimate power which allowed the perpetuation, in each case, of a
regnal community. 16 Some ideological preconditions have already been
sketched: the Carolingian legacy of the preceding century was clearly crucial.
As for Italy: its long-term disunity was further crystallised, as the Carolingian
regnum Italiae became part of the Ottoman empire, though, necessarily, given
often distant rulers, a very distinct part with its own forms of devolved and
indigenous power, while Byzantines, Lombards and Arabs vied for the south.
Thus die three Carolingian successor-states became two; and the tendi century
was also the crucial formative period for the long-lasting linkage of Germany
and Italy.

The continuity of Burgundy’s royal line, acknowledged by odier dynasties
who intermarried with it, enabled diat kingdom to persist diroughout the tenth
century, though more or less dependent on the Ottomans. This was a sub-
Carolingian realm, in which Carolingian traditions and modes of rulership —
tides and charter-diplomatic, signs and symbols, the local church’s ideological
and institutional support - had a continuous history.

A variant explanation is needed when we turn to the lands formerly on the
periphery of the old empire. Here, kings newly emerged from the ranks of
noble warlords were constructing new realms in Leon-Castile, Denmark,
Poland and Hungary. The timing, and the location, of these changes was not
fortuitous: people just beyond the frontiers of powerful states, and hence on
the receiving end of their neighbours’ aggression, had learned to unite in self-
defence, and their leaders to copy something of their adversaries’ methods.
The brunt of the English kingdom’s growing and self-consciously English
power was borne by their neighbours. Both Welsh and Scots responded with a

16 ‘regnal’: Reynolds (1984)^. 254.

Cambridge Histories Online © Cambridge University Press, 2008

Rulers and government

IOI

growing sense of their own identity, and new manifestations of royal authority,
notably in the reigns of Hywel Dda and Constantine II. But, thanks to the
vagaries of dynastic alliance, far-flung water-borne contacts and chance, it was
also through die English kingdom that Carolingian traditions were to be
further diffused to Denmark and Hungary.

Ethnicity played only a limited part in new regnal formations. The world of
ninth-century Latin Christendom had been a world dominated by the Franks.
Even though the Carolingian rulers had not used an ethnic epithet in their offi-
cial titles, theirs was a Frankish empire, won by Frankish arms. Yet it was never
ruled by Frankish laws: rather what constituted its unity was Latin Christianity
maintained by Carolingian power. And it remained polyethnic in legal, and
social, reality. The Frankish identity which survived in social memory through
story and history, and, extended to western Europeans in general, was
acknowledged by outsiders for centuries to come, was more myth than fact.
Carolingian traditions were detachable from it, and could help form the base of
new identities. A ninth-century Carolingian had entitled himself just rex dei
gratia. It was in the tenth century, from 91 1 onwards, that rexFrancorum became
the normal title of west Frankish kings. Though the process was slow, die ruler
of the east Frankish kingdom by the end of the tenth century had lost specific
close association with Franconia, the properly Frankish area of the eastern
realm: the kings of die Franks and Saxons were on the way to becoming
German rulers. Other new political formations acquired ethnic labels in the
tenth century. Rex Anglorum became the usual title of kings who extended their
power northwards from Wessex. In Poland, what had been eight distinct
groupings were replaced, in the mid-tenth century, by the single Polish potestas
of Miesco. Invented ethnicities then as now papered over political cracks: in
association with kingship itself, they too offered a basis for regnal identity - a
basis which could be detached from any particular dynasty, and hence could
bridge dynastic change.

Dynastic strategies, dynastic accidents, were in part responsible for the crys-
tallisation of all those tenth-century realms. Kingship remained embedded in
family structures and family politics. For most kings, die royal family was the
environment in which they were born and grew up. In the kingdoms created
out of the ninth-century Carolingian empire, though by 1000 the direct
Carolingian line had been extinguished in all of them, the assumption
remained that there would be a ruling dynasty; and it was around dynasties that
new kingdoms were being created, in central Europe, Denmark, England and
Wales.

Yet in this same tenth century, kingship became institutionalised in new ways
that showed it outgrowing the family. In the evolving identity of realms, king-
ship came to play an increasingly central role. The distinction between family

Cambridge Histories Online © Cambridge University Press, 2008

102

JANET L. NELSON

resources and regnal resources became less blurred. In the east Frankish
kingdom, the non-Carolingian Conrad I claimed (even if unsuccessfully)
control of regnal lands in Bavaria, far from his own hereditary property. In
west Francia, non-Carolingian kings contrived, in the 920s and again after 987,
to resume regnal lands from Carolingian claimants: indeed in 987, arguably, it
was only thus, by insisting that royal estates were regnal, that Capetian power
could be established at all. In Italy, the payment of regnal dues (the fodruni) was
so well established at Pavia that these continued to be collected by royal officers
throughout the second half of the tenth century: though tire rulers of the
regnum Italiae were frequent absentees, they stayed there often enough, and for
long enough, to maintain plausible claims. Conrad II was depicted by his biog-
rapher early in die eleventh century as having a very clear sense of the distinc-
tion between a private home and the palace as a public building. 17

Two aspects of this recognition of the difference between kingship and
family deserve special attention. First, the royal succession was more con-
sciously and collectively managed. There were pre -mortem arrangements: in
east Francia, Otto I’s son and namesake became co-ruler in 961, co-emperor in
972, while in west Francia, Lothar made his son Louis V co-ruler in 979, and
Hugh Capet, himself consecrated on 3 July 987, had his own son Robert conse-
crated as co-ruler on Christmas Day of the same year. The church absorbed
surplus royals: die illegitimate younger son of Otto I and the illegitimate oldest
son of die west Frankish Lothar. Violence widiin royal families did not cease,
but, when it did occur, was publicly mourned, as in die case of the east
Frankish Thankmar whose killing was disavowed by his half-brother Otto I, or
the Anglo-Saxon Edward, of whose murder a contemporary asserted: ‘no
worse deed than this for the English people was committed since first they
came to Britain’. 18 The ‘people’, that is, die aristocracy, were often recorded as
electors of kings, not only when diere was a change of dynasty, but also on
other occasions including the creation of co-rulers. Nobles thus participated
actively in the management of the succession, and this was occasionally
acknowledged by the draftsmen of royal documents. 19 When noble factions
supported rebellion, they aimed at the restoration of king-centred politics and
dynastic unity. In 9 5 3-4, the supporters of Otto I’s rebellious son Liudolf jus-
tified their action by alleging that the king’s new bride, Adelaide, and his
brother Henry had colluded to monopolise counsel. 20 Insisting on their role —
at once right and responsibility - as advisers of kings, magnates imposed stan-
dards of royal conduct. In the 920s, it was in the name of those standards,
because ‘Charles [nicknamed Simplex, ‘the Straightforward 5 ] listened to and

17 Wipo, Gesta, c. 7, p. 7. 18 Widukind, Res gestae Saxonicae 11, 1 1 , p. 65 ; ASC, s.a. 978, in EHD , p. 2 30.

19 D Lo 4; Sawyer, no. 520. 20 Hrotsvitha, Gesta Ottonis , ed. Winterfeld, lines 73 5—40, p. 225 .

Cambridge Histories Online © Cambridge University Press, 2008

Rulers and government

103

honoured above his leading men (principes) his counsellor Hagano whom he
had picked from among die lesser nobility ( mediocres )’ 21 and, again, ‘hearing drat
Charles planned to summon the Northmen to join him’ against his own
Franks, that diose Franks definitively rejected Charles, chose another king, and
kept Charles locked up. (The magnate who, 300 years later, reminded the
English king Henry III of Charles’ fate was angered, similarly, by royal failure
to be counselled aright. 22 ) The Frankish chronicler noted parallel events in far-
off Italy, where magnates ‘disturbed the realm because of their king’s inso-
lence’. 23 ‘Disturbance’ was a legitimate form of resistance. In west Francia, the
key point was precisely that Charles was replaced, and another king was
chosen. King and magnates needed and sustained each other. By the close of
the tenth century, Abbo of Fleury had summarised that mutual dependence in
a neat pun: ‘since die king on his own is insufficient for all the needs of the
realm, having distributed the burden (onere) amongst die others whom he
believes worthy of honour (honori) he too must be honoured ( honorandus est)
with sincere devotion’. 24 Only a few years later, in 1014, die author of tht Anglo-
Saxon Chronicle strongly implied that loss of magnate support rather than
Danish invasion drove dEthelred into foreign refuge: he returned when ‘sent
for’ by ‘all the councillors’, but on condition that ‘he would govern his kingdom
more justly’. 25

The magnates’ active role as counsellors, hence not just self-appointed but
widely acknowledged representatives, of the realm contributed to a second
important tenth-century change: the tendency for realms to become indivis-
ible. In transjurane Burgundy a divided succession was rejected when Rudolf I
was succeeded by only one of his two sons, Rudolf II, in 912. Three other
similar instances occurred elsewhere soon after: in 919, east Frankish and
Saxon nobles (though not Bavarians) combined to choose Henry the Fowler as
king; in 922, west Frankish nobles acted in similar fashion to replace Charles
the Simple by Robert; and in 924-5, the choice of Mercian and west Saxon
nobles converged on fEthelstan. With hindsight, these events can be seen as
critical for the realms in question: and the near-coincidence was not fortuitous.
Though other options were available, in each case a regnal community at most
three generations old was recognised, and kept in being. Once an undivided
succession had occurred for a further generation or two, division became
increasingly improbable, and eventually undrinkable. If Mercia and Wessex
parted company again, briefly, in 957, that was symptomatic of the persistence
of two regnal communities that long pre-dated tire later ninth century.
Sundered again briefly under pressure of Danish invasion in 1016, the two

21 Flodoard, HREiv, 15, p. 5 77 - 22 Bemont (1884), appendix, p. 341.

23 Flodoard, Attnales, s.a. 922, p. 7. 24 Abbo of Fleury, Liber canonom, PL 139, col 478.

25 slSC, version ‘e’, r.tf.1014, p. 246.

Cambridge Flistories Online © Cambridge University Press, 2008

104 JANET L. NELSON

were reunited a year later in what proved a permanent union of the crowns. By
then, multi-gentile royal titles like ‘king of the Saxons, Mercians and
Northumbrians’ had fallen from use in charters (though this one survived in
some ordines manuscripts), to be uniformly and definitively replaced by ‘king of
the English’.

It was not that anyone formulated a principle of indivisibility ( Unteilbarkeit )
or primogeniture. Dynastic accident played a part, allowing unity to persist
long enough for custom to congeal. In the east Frankish realm (and also the
Italian one) Otto I, then Otto II, died leaving just one legitimate male heir. In
west Francia, however, the choice of Hugh Capet in 987 was not imposed by
lack of a Carolingian alternative: Charles of Lorraine was deliberately rejected.
According to Richer, the archbishop of Rheims recommended Hugh to the
magnates as one ‘whom you will find a protector of private interests as well as
of the public interest’. 26 The dwindling of royal resources had already meant
that separate subkingdoms for younger brothers could no longer be funded:
hence die dangerous poverty of Charles of Lorraine. In England, the fact that
dEthelstan remained unmarried, hence heirless, could suggest a family pact
whereby an older man ruled as a kind of stake -holder for younger half-broth-
ers: fraternal succession josded with filial succession in tenth-century England.
Here as elsewhere younger brothers or nephews (and their potential support-
ers) reconciled themselves, more or less willingly, to exclusion: the several sons
of fEthelred II were not endowed with subkingdoms in their father’s lifetime
(the final, desperate, bid by Edmund Ironside in 1016 was a unique exception).
Successions were disputed by would-be rulers of whole realms, with division
increasingly infrequent. In the kingdom of Leon-Castile, as in England, there
were several cases of fraternal succession in the tenth century. The nine-year
regency (966-75) of Elvira, daughter of Ramiro II and abbess of the royal
house-convent of San Salvador in the city of Leon, for her young nephew
Ramiro III (born 961), the first royal minority in Leonese history, indicates a
new stability of the realm. Here as elsewhere, the aristocracy’s role is better
attributed, positively, to support for the dynasty than, negatively, to egoistic
hopes of exploiting a royal minority. Where nobles collectively continued to
invest in a regnal tradition, that realm tended to survive as a unit; and, where
charters are available to document this, nobles’ titles and self-presentation
suggest that they considered themselves kings’ collaborators, subordinates and
surrogates. Even if they were normally based far away, they were potentially
involved in face-to-face contacts with kings. Their exercise of lordship was in
principle legitimated by being part of the realm. There was no alternative eccle-
siastical legitimation for nobles: they remained, as Richard Southern observed

26 Richer, Mistoriae iv, 1 1 , p. 162: ‘non solum rei publicae sed et privatarum rerum tutorem invenietis’.

Cambridge Histories Online © Cambridge University Press, 2008

Rulers and government i o 5

of west Frankish counts, ‘shockingly unconsecrated and dumb’, 27 lacking, that
is, theoreticians or apologists: the church called down heavenly blessings upon
kings but provided no separate prayers for nobles. The evidence for the devel-
opment of princely accession-ceremonies of various kinds (though never
including anointing) post-dates die tenth century. 28

In die lands beyond the Carolingian empire, other factors helped determine
whether a kingdom materialised and/or persisted. England’s formation is
partly explicable in terms of die familiar dynamic of acquisition and reward
that enabled the west Saxon kings, not without setbacks, to mount a kind of
imperial expansion against the Danelaw and at the same time consolidate
support in English England. Even the resultant composite realm was relatively
small compared with France and Germany; and smallness allowed that greater
degree of cohesion which continues to impress modern English historians.
Another important factor was linguistic community: despite regional varia-
tions, Old English could be understood throughout the realm, the more readily
once efforts had been made in the second half of the tenth century to standar-
dise tire written vernacular.

How did the Ottomans manage in a realm so very much larger? Because
both date and place are recorded in their charters, the Ottomans’ itineraries can
be reconstructed. There were diree core regions: the Harz area of Saxony
(where in the 950s, happily for the Ottomans, silver lodes were discovered),
lower Lotharingia around Aachen, and the Rhine-Main area (with key bases at
Ingelheim and at Frankfurt where Henry’s son Otto I stayed often). In holding
onto their palaces and estates in the latter two regions, die Ottomans continued
Carolingian patterns. In their relentless journeyings, the Ottomans signalled,
and exploited, die dual basis of their regime and its legitimacy: they were ‘kings
of the Saxons and the Franks’. At home in Saxony itself, they preferred the
monasteries and convents staffed by their own, and the aristocracy’s, kin. Two
convents, Quedlinburg and Gandersheim, had special importance, both with
Ottoman abbesses who became icons of die royal house in their own right.
Between them these places performed the key functions of dynastic centres:
guardianship of the tomb of the dynasty’s founder, keeping of collective
memories of royal deeds, providing of frequent hospitality for the court,
hosting of great assemblies, and last but not least praying for die prosperity of
king and realm. These indeed provided the ritual centre of the kingdom. The
Ottomans were able to draw into it the powerful from the various regna of their
realm so that the net of intermittent contacts spread wide. Churchmen kept
the net in repair. In their minds the realm was a res publica — and the distinction
between that and private interests remained. Abbot John of Saint- Arnulf, Metz

27 Southern (1953), p. 99. 28 Hoffmann (1962).

Cambridge Histories Online © Cambridge University Press, 2008

106 JANET L. NELSON

depicted his hero and namesake John of Gorze as ready to leave a rustic
‘private life’ for the twin challenges of involvement in spiritual reform and in
royal service. 29 Thus Carolingian habits of thinking proved resilient: they sus-
tained, at any rate for churchmen, habits of acting that took John of Gorze
from country to court and, eventually, far beyond the frontiers of the Reich.

What made the little kingdoms viable, still, was not only their adequate eco-
nomic base in land and cattle but their rootedness in law and tradition. In
Ireland, where geography might seem to have predestined an island-state, the
past continued to dominate the present: despite external threats, no High
Kingship was established covering the whole island. Regional overkingships
emerged fitfully: four eventually survived (Tara, Cashel, Connacht and
Leinster), but several others proved shortlived. More important, many little
kingdoms, tuatha only 1 6-20 kilometres in radius, survived as well, with kings
unsubordinated to any higher-level ruler. In the old Carolingian lands, transju-
rane Burgundy which became a kingdom late in the ninth century outlasted the
tenth, partly because geography protected it, still more because of dynastic
continuity. Brittany, on the other hand, failed to survive as a realm into the
tenth century. Smallness of scale was not the only, or most, important reason:
dynastic disputes at a critical moment inhibited die formation of identity
around an indigenous kingship, while Viking attacks crippled the ruler’s
resources. Still more fundamentally, political traditions, and not least ecclesias-
tical ones keeping all the Breton sees suffragans of Tours, situated Brittany
firmly within the west Frankish realm.

Elsewhere, nascent kingdoms were able to maintain and expand their terri-
torial range, and at die same time to integrate, more or less effectively, dieir
internal spaces. Though scarcely anything can be known of royal itineraries,
efforts to establish royal cult-centres can be inferred. In Scotland, under
Constantine II, St Andrews became such a centre, representing a substantial
eastwards shift in focus and a determined appropriation of the resources of
the former Pictish kingdom. Glastonbury, in die ancient heardands and ‘more
princely part’ 30 of Wessex, especially after Edgar’s burial there, bid fair to
establish its claims to that role in England. In the 980s, the chain of Harald’s
fortresses in Denmark linked old and new centres, with a pronounced shift
eastwards towards the resources of southern Sweden. Poland and Hungary
evinced a similar process of political centring with die choice of sites for the
new bishoprics of Gniezno (1000) and Gran (1001). Still-migratory monarchs
were nevertheless conscious of where the seat of the realm lay.

29 Vita Iohannis Gor%iensis, c. 51, p. 351. 30 Asset, Ufe of Kingsilfred, c. 12, ed. Stevenson, p. 10.

Cambridge Histories Online © Cambridge University Press, 2008

Rulers and government 1 07

ROYAL RITUALS

In the tenth century, the king-making rites pioneered by Frankish clergy in
Carolingian times became more or less standardised and generalised through
much of Latin Christendom, acquiring permanent form as liturgical ordines.
These rituals were specific to kingship, marking, and demarcating, the unique-
ness of monarchy. The role of clergy in elaborating such rituals was crucial, but
laity played their part too. Even in the ordines themselves, the aristocracy’s role is
evident, as witnesses and primary audience, even as active participants in the
acclamation and enthronement of the new king. While clergy now stage-
managed and took leading roles in proceedings inside the church, lay leaders
were conspicuous in other rituals performed outside the church. Principes
acknowledged the new ruler in a form of election preceding the liturgical rite.
They served, and joined in consuming, the food and drink at the banquet fol-
lowing the clerical proceedings. They accompanied the new king on the
journey through his realm in which he presented himself, symbolically and
practically, to his new subjects. The ordines thus give only a partial picture. It is
the historical writers, and especially Widukind and Byrhtferth, who reveal
something of the full and lengthy process whereby kings were made and dis-
played. Together, the two very different types of source-material, history and
liturgy, throw light on one important way in which kingdoms took shape as
political communities in this period.

The clericalising of a substantial part of the process did not happen uni-
formly, or necessarily in linear progression. The Carolingian realms of the
eighth and ninth centuries were diverse in substructure, and underwent varying
experiences. This diversity in formation helps account for important differ-
ences in the tenth century, especially between east and west Francia. In the
post-843 east Frankish kingdom, it is possible that no king was consecrated by
local clergy before Conrad I in 91 1 (Charles the Fat apparently had himself
consecrated in 880 by the pope). In any event, when Henry the Fowler in 919
decided to forgo anointing by the archbishop of Mainz, he was not breaking
with long-established tradition. Modern historians have argued that accep-
tance of consecration made a king-elect in a sense behoven to his main conse-
crator, and it is true that Henry mistrusted the archbishop. It is unlikely,
though, that Henry was driven by anxieties about clerical control. Complex dis-
cussions with lay magnates probably account for die five-month delay between
his predecessor’s deadi in December 918 and his ‘designation’ in May 919 by
‘the whole people of the Franks and Saxons’. 31 He may have wanted to display
to his key supporters in Saxony his independence of Franconian domination.
Perhaps still more important, given the strength of the component regna within

31 Widukind, Res gestae Saxonicae i, 26, p. 34.

Cambridge Histories Online © Cambridge University Press, 2008

108 JANET L. NELSON

his realm, he needed to reassure the greatest magnates (duces) of those odier
regna that he meant to respect what they saw as dieir customary rights, and
would be committed to a consensual style of government: this was what they
were to understand by Henry’s friendship (amicitid). The prestige of Henry’s
monarchy grew. His successor was not only acclaimed by duces and milites
outside Charlemagne’s church at Aachen, but also consecrated inside it . 32 A
generation later, an elaborate east Frankish royal ordo (formed in part from
west Frankish ingredients) was included in the Romano-Germanic Pontifical
copied at Mainz and widely diffused . 33 According to this rite, the designatus prin-
ceps divests himself of cloak and weapons, and prostrates himself in front of
the altar steps. Spiritually abased, symbolically annihilated, he rises to assume a
new persona. He swears to ‘rule and defend his people according to the
custom of their forefathers’, and clergy and people acclaim him. The anointing
signals the transformation of princeps into rex, making him strong against his
enemies and capable of ruling his people justly. There follow his re -arming,
and reinvestiture with arm-rings, cloak and sceptre, his coronation, and his
enthronement whereby he is raised up in the manner of David and Solomon,
‘to be king over God’s people’. This ordo was used for Otto Fs successors. By
the end of tire tenth century, Ottoman monarchy had reached new heights of
exalted representation and display, both in the private form of book-illumina-
tion and in the rather public form of table-manners at court, with die ruler
dining alone and on high. Otto III was of course an emperor, and imitating
Byzantium . 34 By c. 1000, diough no other ruler in the west would or could
match Otto’s pretensions, rites of royal consecration were in regular use in
Spain, France and England; and even where they were not, as in Scodand or
Denmark, Christian glosses were starting to be superimposed on indigenous
rulership. Mentions of queen-makings, often linked with marriage, became
more frequent in narrative sources, and pontificals began to offer queenly
ordiner. in one widely diffused rite, die prayer accompanying investiture with the
ring assigned the queen responsibility for spreading knowledge of the faidi
among barbarian peoples . 35

Royal funerals, relatively unremarked earlier, become better documented in
die tenth century. In the old Carolingian realms, the explanation is not that
diese rites had been newly clericalised: rather, increased attention was being
focused on them by kings, but also by leading clergy and laymen. In 973, Otto
Fs funeral became the occasion for a re-enactment of his son’s election ‘ab
integro, by the whole people’, and the electus then bore the old emperor’s body to

32 Widukind, Res gestae Saxonicae n, i , p. 5 5 .

33 Pontifical Romano-Gertnanique , ed. Vogel and Elze, 1, 246—5 9.

34 Thietmar, Chronicon iv, 47, p. 1 84.

35 ‘Erdmann’ Ordo , prayer ‘Accipe anulum’: Schramm (1968), p. 221.

Cambridge Histories Online © Cambridge University Press, 2008

Rulers and government 109

Magdeburg for burial. 36 The west Frankish king Lothar in 986 was given ‘a
magnificent royal funeral’ at Rheims, his corpse, clad in gold and purple,
carried on a special royal bed by the leading men {primates) of die regna (the
plural enhances die imperial stature of the kingdom - regnum— to which, a few
lines further on, his son succeeds), with the regalia and especially the ‘gleaming
crown’ prominendy on display in die procession, followed by his retinue. 37
Royal mausolea were numinous places, as Otto III knew when he opened
Charlemagne’s tomb at Aachen in 1000, and publicised the link between his
own ‘renewed’ empire and his predecessor’s. Further north, a no less political
message was purveyed when Harald Bluetooth erected a massive runestone at
Jelling to commemorate his royal parents and at die same time proclaim his
own rulership over all Denmark.

It was in die tenth century that the three Magi of the Gospel story were
recast as kings, and began to be depicted with crowns. 38 The crown had now
become the royal attribute; crown-wearings became great ritual occasions for
the Ottomans particularly, and perhaps for odier contemporary rulers in west
Francia and England too. Sets of regalia, crowns foremost among them,
acquired a further significance when power was being passed on not just from
one ruler to the next but from one family to another. Widukind describes die
dying Conrad sending to Henry I ‘these insignia, the holy lance, the golden
arm-rings, the cloak, die sword of the kings of old, and the crown’. 39 The lance
here is anachronistic: it was only acquired for die Ottoman royal treasury by
Henry himself (probably in 926) but thereafter became the kingdom’s chief
talisman of military success. Later in the century kings of Poland and Hungary
would seek to acquire holy lances of their own. The regalia, and especially
crowns, contributed to a depersonalising of kingship as, transmitted through
generations and across dynasties, they became symbols of state. They were ina-
lienable perquisites of the royal office, and they were (or in the case of the
lance, swiftly became) specific to particular kingdoms. Even more homely
items of royal attire acquired a unique quality. Those same west Frankish mag-
nates who resented Hagano’s hold on Charles the Simple objected particularly
to the royal favourite’s frequendy and in public taking die cap off die king’s
head and putting it on his own. 40 Otto II’s royal banner was publicly displayed
before batde. 41 Special rituals of reception and farewell marked the comings
and goings of kings. Royal diplomata worked by being multiplied and distrib-
uted: the documents themselves, often large and splendidly produced, with
seals showing the ruler in majesty, were substantial communicators of regality.

Other royal rituals, feasts and fasts, celebrations of peace and friendship and

36 Widukind, Res gestae Saxonicae in, 76, p. 1 27. 37 Richer, Historiae hi, 1 10, p. 142.

38 Sacramentarium Fuldense saeculi x, plate 16. 39 Widukind, Res gestae Saxonicae 1, 2 5 , p. 33.

40 Richer, Historiae 1, 1 5, p. 38. 41 Fichtenau (1984), pp. 71—2 (1991, p. 48).

Cambridge Histories Online © Cambridge University Press, 2008

I IO

JANET L. NELSON

alliance, displays of largesse and clemency and religious devotion, could be
imitated by magnates and by lesser nobles. The availability of an acknowledged
ritual repertoire seems to have stimulated inventiveness on the part of laymen
as well as clergy: the tenth century was an age not just of ecclesiastically con-
structed liturgy but of proliferating sign and gesture in a variety of social con-
texts. These were forms of communication that transcended linguistic barriers
and could be received and responded to simultaneously in multiple ways.
Rituals of supplication as performed by faithful men (fideles) before kings were
replicated by faithful men and fighting men (milites) before their lords. Ritual
thus associated the elite’s status and functions with those of the king. Outsiders
who learnt the rules could qualify as insiders: Miesco of Poland never
remained seated in die presence of Otto I’s representative Margrave Hodo, 42
and when he attended Otto’s assembly, he presented the young emperor with a
camel for his menagerie. Royal zoos were another form of monarchic self-rep-
resentation, derived in this case from antique and Byzantine models. Most
important of all, Miesco had become a Christian and showed himself willing
to play by the most fundamental of rules. But ritual was not always about
togetherness. It could also assert distance, as when Otto summoned the
defeated Count Eberhard of Franconia and his leading followers ( principes
milituni) to Magdeburg in 937. Their offence was not only rebellion, but the
destruction of a stronghold which had belonged to one of Otto’s Saxon trus-
ties, Bruning. The count had to present the king with 100 talents’ worth of
horses, and his men had to perform the ritual humiliation of carrying dogs.
This was political theatre indeed, necessarily performed in public, before
Otto’s new court (his reign was scarcely a year old) consisting of its essential
components: Saxons and Franks. 43 The echo of the Carolingian harmscar , 44 a
similar ritual punishment, was presumably not lost on such an audience.

Rituals distanced kings from their subjects; rituals closed the gap between
them. Kings did not choose between these as alternatives, either cementing
loyalty through horizontal bonds of conviviality and amicitia, or vertically
asserting their own superiority. Friendliness and lordliness were regarded as
perfectly compatible; and effective kings operated in both registers simultane-
ously. Alfred of Wessex, in the 890s, recommended a ‘friendship’ which bound
the receiver of wealth to the lord who gave it: should one’s lord call for active
service, ‘it is better to renounce the gift and follow the giver, who acts as the
guardian both of the wealth and of his friendship’. 45 In the Frankish world,
such relations of political friendship, though they continued Carolingian prac-

42 Thietmar, Chroniconw , io,p. 232. 43 Widukind, Res gestae Saxonicae n, 6, p. 61.

44 MGH Capitularia regum Francorum , index, sv.

45 Alfred’s version of Augustine’s Soliloquies , trans. Keynes and Lapidge, p. 141 .

Cambridge Histories Online © Cambridge University Press, 2008

Rulers and government

rice (Charles the Bald had established amicitia with Bernard of Septimania in
841, and Charlemagne had surrounded himself widi amici in baths and
bedroom), are more frequendy mentioned, and more evidently clodied with
ritual, in die tendi century. In the early years of his reign Henry die Fowler was
in urgent need of friends to recognise and support his kingship. In 921, he and
Charles die Simple swore friendship as between ‘eastern’ and ‘western’ kings. 46
Two years later Henry joined Gislebert of Lotharingia to himself in friendship
and gave him his daughter in marriage. 47 For contemporaries, context
explained, and ritual clarified, the different infiections of the same word: the
elaborate arrangements of 921, which culminated in the two kings’ meeting on
a ship anchored in the middle of the Rhine, expressed equivalence, whereas in
923 friendship between king and non-king, senior and junior, demonstrated
Henry’s ‘liberality’ towards Gislebert, and marked distance between them. A
ritual reception similarly displayed hierarchy at die same time as it forged bonds.
In 924, King Radulf awaited Duke William of Aquitaine on the nordiern bank
of the Loire (as Charles the Bald had awaited Aquitanian magnates in the ninth
century): it was William who had to make the river-crossing and, when he met
the king, leap down from his horse and approach on foot while Radulf
remained mounted. Only then did Radulf offer him a kiss, which signified
peace. But peace, as in diis case, could restate, rather than elide, distance. The
outcome was what Radulf wanted: William ‘committed himself to the king’. 48
Kings maintained their prestige and uniqueness in die midst of ritual prolifera-
tion. Lay aristocrats could have seals, but the iconography of royal seals
remained distinct: only the king was shown seated in majesty. In west Francia, at
any rate, he was addressed as Your Majesty. Rituals of friendship found their
context in representations of divinely blessed monarchy. The special royal bed
on which the dead west Frankish king was laid 49 recalled the uniqueness of the
living king. No-one else might lie with impunity in the bed prepared for the
king: when a Saxon magnate did diis, it was probably meant and understood as a
ritual of rebellion. 50 The disposition of kings’ daughters made the same point
in another way. They were (Henry I’s daughter Gerberga was a rare exception)
given in marriage to foreign princes, or placed in royally supported convents.
Subjects did not share their beds as spouses. Great aristocrats, by contrast, reg-
ularly married their daughters into other magnate families to produce die many-
stranded patterns of alliance visible in tenth-century genealogies.

The cults of saints linked kings and aristocrats. In east Francia, groups of

46 Constitutiones et acta publica , no. i , p. i ; Altholf (1992).

47 Widukind, Res gestae Saxonicae 1, 30, p. 43: ‘[rex] liberaliter eum coepit habere . . . affinitate pariter cum
amicitia iunxit eum sibi . . 48 Flodoard, Annales, s.a. 924, pp. 19—20.

49 Richer, Historiae in, 1 10, p. 142.

50 Thietmar, Chronicon 11, 28, as interpreted by Leyser (1994), pp. 198— 201.

Cambridge Histories Online © Cambridge University Press, 2008

I 12

JANET L. NELSON

royal and noble names entered in the Memorial Books of St Gallen and
Reichenau preserve evidence of particular occasions of collective devotion,
notably in the reign of Henry the Fowler. Kings and nobles alike had their family
saints, their relics, their mausolea. Ingeneral, though, even the greatest magnates
could not equal kings in the quantity and quality of their relic-collections and
saint-patrons; and the Ottomans too would set about gaining that edge once
firmly in power. For his son, Henry the Fowler gladly accepted an English bride
of St Oswald’s ‘holy line ’ 51 who would help compensate for the Ottomans’ own
distinctly un-royal antecedents. More important to Otto I than Oswald was
Maurice, a soldier-saint specialising in victory, but whose relics, unlike Oswald’s,
were in Otto’s possession. They strengthened his men’s morale and so Otto’s
own prestige. No tenth-century king was a keener relic-collector thandEthelstan:
‘we know you value relics more than earthly treasure’, wrote the abbot of Saint-
Samson, Dol, in a covering letter for gifts in point; and Duke Hugh of Francia
sent him StMaurice’s standard. 52 dEthelstan may have seen these acquisitions early
in his reign as prerequisites for future military successes. Right relations with the
saints entailed giving as well as receiving. On campaign against the Scots,
dEthelstanpaused at Cuthbert’s shrine at Chester-le-Street, and offered lavish gifts
to the saint. Such rituals of largesse and devotion at sites of supernatural power,
from Chester-le-Street to Winchester and from Exeter to Canterbury, enhanced
royal authority and underpinned a newly united imperial realm.

ROYAL GOVERNMENT

The west Frankish kingdom has been treated in much recent historiography as
a tenth-century paradigm. ‘France’, allegedly, was ‘without a state’, and in its
vast ‘kingless territories’ such as Aquitaine there was a falling-back on ‘order
and social obligations transmitted from the period preceding statehood ’. 53
Generalise from the French model and you have a tenth century for which talk
of royal government would seem redundant. France’s typicality can be dis-
puted, however: historians of Denmark, Poland, Spain, Germany and England
represent the tenth century as a period of strenuous royal activity and even
some institutional growth. The notion of ‘government’ can be variously
defined. It was just a generation after the year 1000 that Duke William of
Aquitaine urgently consulted King Robert II over a portent, the falling of a
rain of blood, that had affected Aquitaine, while Duke Richard II of
Normandy sent a warning to the same king about heretics at Orleans . 54 Both

51 Hrotsvitha, Gesta Ottonis , ed. Winterfeld, lines 95—6, p. 207. 52 EHD , no. 228, p. 892.

53 Fichtenau (1991), p. 423, and cf. p. 391.

54 The king reported Duke William’s request to Abbot Gauzlin of Fleury, Fulbert, ep. 2 (appendix b),
pp. 274/6.

Cambridge Histories Online © Cambridge University Press, 2008

Rulers and government

1 1 3

dukes still had a sense of the king’s uniqueness, and of his responsibility for die
religious well-being of the whole kingdom. The king responded with energy to
both demands - in the one case demanding learned scrutiny of appropriate
books, in the other, arresting and burning the heretics. How had such percep-
tions survived a century when royal government was allegedly non-existent in
west Francia? One answer lies in an exceptionally rich Carolingian legacy,
including precociously developed rituals of royalty as well as more long-
standing and deep-rooted sub-Roman forms and traditions of monarchy. Are
we then to imagine tenth-century west Francia as resembling Clifford Geertz’s
vision of nineteenth-century Bali where ‘power served pomp’ - diat is, where
the state existed merely as a ceremonial display whose sole and solipsistic func-
tion was to maintain itself? West Frankish kingship was not as pompous as all
that. Moreover, it surely functioned, and not least in its more theatrical
moments, for others too. Because it, and the idea of die realm’s geographical
integrity, were so well entrenched in the minds of its leading subjects, because
aristocrats were so habituated to drinking of their own power in terms of dele-
gated royal power, because churchmen were so imbued with ideals of service
to die realm, because social memory was so deeply impregnated by written
records as well as oral tales and poems in which royal deeds and judgements
were central themes, tenth-century monarchy, despite chronic material weak-
ness and in die case of some kings serious loss of prestige, retained legitimising
authority. If centralised judicial institutions - a supreme court - are the hall-
mark of the state, diere was no state here. Yet royal authority might generate a
kind of power as the imagined guarantor of order and focus of fidelity.

In the 930s, the last heir to the independent kingdom of Provence, die ambi-
tiously named Charles-Constantine, ended his days as a count, the faithful man
of the west Frankish king Radulf. Faithfulness to the king still meant some-
thing. For the king was no cipher. In Flodoard’s Annals, just one year (931) in
the life of King Radulf shows him incessantly active: he went from Burgundy
to Vienne to accept the submission of count Charles-Constantine, then to
Saint-Martin, Tours ‘for the sake of prayer’, then back ‘into Francia’ to sup-
press rebellion, besieging two strongholds, and making a truce; he ‘sent letters
to the clergy and people of Rheims about die archiepiscopal election’, moved
to Attigny, engaged in furdier negotiations with the rebels, and headed off the
rebels’ potential ally, the east Frankish king Henry, by sending hostages; he dien
captured Rheims after a three-week siege and installed his candidate as arch-
bishop, imprisoned the rebel bishop of Chalons and installed ‘his own cleric’ in
the see; he besieged the rebel leader in Laon and captured it; and so back to
Burgundy and the recovery of more castles from rebels. 55 Here then was a king

55 Flodoard, A.nnales , s.a. 931, pp. 46— 5 2 .

Cambridge Histories Online © Cambridge University Press, 2008

JANET L. NELSON

114

beset, his control of tire heartlands and key churches of Francia direatened,
even Rheims and Laon temporarily lost. Yet he rallied, showed diplomatic and
military skills, reasserted his control of key churches, recovered the royal seat
(secies) of Laon, preserved for his successors something of dignitas - royal
status. What is missing in all this frenetic activity is the exercise of jurisdiction.
Radulf and his successors, unlike their contemporaries in east Francia and
England, were unable to impose judgement upon powerful subjects. The west
Frankish king simply could not command sufficient resources, could not
muster enough troops, to act in the political arena as more than one regional
prince (and in military terms a second-rank one) among others. The one trump
card he might still play was his kingliness: he could appeal to other kings for
help against rebels - as Lous IV did twice to /F.thelstan, and many times to
Otto I, both of them relatives as well as fellow-kings.

Power in the tenth-century west Frankish kingdom was devolved and
regionalised. This was hardly new. A Golden Age of Carolingian justice has
been presented in some recent historiography, and the Peace movement por-
trayed as a response to post-Carolingian ‘crisis’. Yet consuetudines, in the sense of
seigneurial power, had already been mentioned in die edict of Pitres 864, and
so too had castles (castelld) built ‘to oppress the locals’ and apparently doing drat
quite effectively even without mottes. 56 Castles did not perse generate change or
signify a new phenomenon of seigneurial power. Rather, dieir effect depended
on the diffusion of building technology and on political context; on what other
kinds of power existed, in the vicinity, or at a higher level (as when Charles the
Bald had asserted his right to audiorise fortifications: we have no idea with
what success). The same was true of the markets that proliferated in the ninth
and tenth centuries: rulers might claim a cut of tire profits from commercial
exchange, but local aristocrats would stake their claims too. Outcomes would
depend on how far die king was able to impose his power in a given region. The
notion of a ‘degradation’ of public to private, and of ‘crisis’ c. 1000, oversim-
plifies complex processes. If royal power was so weak throughout the tenth
century, it becomes hard to explain why the ‘crisis’ was so long delayed. No one
generalisation works for the west Frankish kingdom, less still for the whole of
the west. The chronology and the geography of change remain obstinately
diverse.

Useful comparisons with west Francia are offered by Denmark, Poland and
England: new realms based on military force which borrowed the clodies of
Carolingian kingship, where ideological superstructure was attached (in what
Marx saw as normal fashion) to material base, pomp serving power. In west
Francia, oldest of the old realms, die ruler wore the clothes of majesty; yet thus

56 Tniustas consuetudines noviter institutas.’; ‘castella et firmitates . . . sine nostro verbo . . . disfactas

habeant’, MGH Capitularia regum Francorum n, 322, no. 272, c. 28; n, 328, additional c. 1 .

Cambridge Histories Online © Cambridge University Press, 2008

Rulers and government

1 1 5

attired, he could deputise for an absent state. The capital of kingship had dwin-
dled: nevertheless something of its political credit remained. It could not
coerce, but it could sometimes impress. This was true, though with a less strik-
ing disjuncture between ideal and realities, of other tenth-century realms as
well. Kings, in west Francia and elsewhere, worked immensely hard to maintain
their credit- worthiness.

Late Anglo-Saxon England is sometimes depicted as such a strong central-
ised state that it seems more Carolingian than the Carolingians. Before its
uniqueness is too enthusiastically accepted (and then celebrated), some dis-
counting has to be done for the centripetal thrust of surviving evidence: while
there is an absence here of the kind of thick description supplied elsewhere by
chroniclers like Widukind or Richer with their revelations of full-blooded and
untidy incident, there is relative abundance of legal material which tends to
exaggerate the statelike appearance of tire tenth-century realm. Uniquely in
this period, Anglo-Saxon England presents a stream of royal legislation. The
coinage, meticulously studied, has been proved to be centrally coordinated.
Equally well-studied royal charters have revealed the range, and precision, of
royal largesse. Scholars working on the laws and the coinage, and also back-
wards from Domesday Book, have been able to identify agents and agencies of
the state. The king might direct ealdormen to publish and implement new leg-
islation, like ninth-century Carolingian counts . 37 Charter witness-lists show
that up to half a dozen ealdormen commonly attended the king. Some ealdor-
men were west Saxons exported to positions of virtual viceroys in the acquired
kingdoms of Mercia, east Anglia and Northumbria; others were local magnates
who threw in their lot with west Saxon kings. Ealdormen of both types
married into the royal family. Kings exercised some control over appointments,
and so qualified the tendency towards hereditary office - qualified only up to a
point, for men with hereditary claims might well be chosen precisely because
they tended to have the local clout that made them effective office-holders.
Prescriptive evidence has reeves and port-reeves supervising a range of royal
resources in countryside and towns. Income from customs levied on trade at
London would, if actually raised, have been far more valuable, and more pre-
dictable, than Welsh tribute. Moneyers working to central specifications pro-
duced silver coins of standard type, weight and fineness; and the locations of
mints and markets were royally authorised. Territorial administration, based on
shires, was extended from Wessex northwards. Landowners brought their dis-
putes before shire-courts in the presence of shire-reeves (sheriffs). Fines for
crimes were paid to the king. Compared with other contemporary rulers, the
kings of England can be shown to have had a large cash income. In the late

57 iv Edgar, c. 1 5.1 ,EHD y no. 41, p. 437.

Cambridge Histories Online © Cambridge University Press, 2008

JANET L. NELSON

1 16

tenth and early eleventh centuries, tire king and his advisers were able to organ-
ise the payment of large sums of money as tribute to Scandinavian attackers.
Wills and other private documents show royal interests operating through local
judicial institutions on a day-to-day basis: ‘shire courts were royal courts’. 58

Nevertheless, these same wills and documents show a fairly rampant and
well-entrenched aristocracy, ecclesiastical as well as lay, sharing in the tasks and
profits of regional goverment. The famous Fonthill case of c. 900 is just one of
the best examples of magnate power operating through, and skewing, royal
courts and legal procedures. Armies were recruited from aristocrats and their
followings. Towns, till recently regarded as bastions of royal control, can be
seen as, at the same time, centres of aristocratic power. Urban development,
the coinage and the fiscal system funnelled wealth into the hands of magnates
and town-dwellers as well as (perhaps, cumulatively, more than) kings. Though,
thanks to Edgar’s and fElfthryth’s interest in ecclesiastical reform, the king and
queen wielded heavy influence over the church monastic and secular, leading
churchmen were often powerful local aristocrats, who could ease their
kinsmen into lucrative posts as reeves and lease-holders of church land. The
Ely evidence shows Bishop fEthelwold of Winchester doing something similar
to extend his territorial power in east Anglia. 59 Into the picture of a church
integrated particularly closely with the state, involved in organising the perfor-
mance of military service, systematically praying for the welfare of king and
realm, and offering personnel and resources as agents and instruments of royal
government, there need to be set the provincial prince-bishops of the Tenth-
Century Reform.

Moreover, royal power was very unevenly distributed through the realm.
England, composed of formerly independent kingdoms, remained an agglom-
eration of provinces. In the midlands and the north, magnates ruled tire roost -
one famous example was actually nicknamed ‘Half-King’. The reach of inten-
sive royal government was limited effectively to the south - old Wessex, where
the king’s lands and residences were concentrated (and even here with some
notable blanks). Further north, and especially northwards of a line drawn from
Chester to tire Wash, the evidence of royal action thins dramatically: there were
no mints except for York, no royal monasteries, few royal stays. Here kings
operated through occasional symbolic interventions, often following up mili-
tary success, rather than day-to-day involvement. When dEthelstan granted a
vast tract of die north-west to the archbishop of York, 60 he was inviting a pow-
erful, but far from reliable, local potentate to make what he could of some
tenuous claims to ecclesiastical jurisdiction in lands drat had recendy attracted
many Scandinavian immigrants: so, from the king’s point of view, a political

58 Wormald (1986), p. 162. 59 Liber Eliensis, ed. Blake, book n, pp. 63—236.

60 Sawyer, no. 407, EHD , no. 104, pp. 548— 51 .

Cambridge Histories Online © Cambridge University Press, 2008

Rulers and government

1 1 7

gesture rather than a sign of prior control. Not long after Eadwig burned
Ripon in 948, a west Saxon clerk now attached to St Cuthbert’s community at
Chester-le-Street took the opportunity to add west Saxon glosses to that prime
Cuthbertian relic, the Lindisfarne Gospels, thus appropriating it for Wessex
while the book itself remained in its Northumbrian home. Kings based south
of the Thames had no option: they relied on ruling at a distance, making exten-
sive use of allies and supporters among the aristocracy of tire midlands
(Mercia) and the north. Occasionally, these men’s attestations of royal charters
show them in attendance on the king - often enough, though, to suggest their
acceptance of his lordship, and the benefits of attachment thereto.

Thus, if in one sense ‘shire courts were royal courts’, they were at the same
time, and perhaps more routinely (though routine evidence is rare), aristocratic
ones. In a famous Herefordshire case, though the shire-reeve was present at
the hearing, the proceedings were managed and the outcome was apparently
sewn up by the beneficiary, Thorkil, die local strong man who could sway the
shire community and also happened to be in favour with the king. 61 In another
famous case, that of Wulfbald, what from ‘die official point of view’ 62 looked
like crimes, and were written up as such after the event, one-sidedly, could
equally well be seen as aristocratic self-help in pursuit of ‘a family dispute
about inheritance’ and one which, significantly, turned on a woman’s claims, for
these were the kind that frequently aroused contention and allowed kings to
intervene in the name of protecting the weak. 63 Ravaging and murder may well
have been usual ways of pursuing such claims in late tenth-century England
(‘harrying’ is mentioned more than once as a royal method for disciplining the
recalcitrant 64 ) . That the king in this case got not one but two large assemblies to
support the forfeiture of the property (which the king then gave to his mother)
suggests that the king was pushing into a controversial area of property law -
indeed pushing his luck. It was the sort of royal action which, pursued in Salian
Saxony, would evoke cries of ‘calumny’ and fierce resistance in the name of
custom. The hard cases of fEthelred’s reign show politics inseparably entan-
gled in the law. External opposition in the shape of Scandinavian attackers pro-
vided a focus for the disgruntled and the dispossessed. Was dEthelred’s
government strong or weak? The terms on which, after his flight to Normandy,
the king was ‘sent for’ by ‘all the councillors’ suggest that for those on the
receiving end, strength might be hard to distinguish from injustice. 65

Tenth-century Welshmen no doubt had varied views of English strong
government. Welsh kings began to establish a more assertive kind of royal
authority of their own, notably in the kingdoms of Gwynedd and Dyfed. A

61 Sawyer, no. 1462, EHD , no. 135, pp. 602—3. 62 Keynes (1 991), p. 79.

63 Sawyer, no. 877. 64 ASC \ s.a. 1000, 1014, pp. 237, 247.

65 ASC \ version ‘e’, s.a. 1014, p. 246; cf. above, p. 103.

Cambridge Histories Online © Cambridge University Press, 2008

1 1 8 JANET L. NELSON

collection of legal tracts is traditionally associated with Hywel Dda (d. 949 or
950) who ruled over both those kingdoms. Coins were issued at Chester in
Hywel’s name. Son-to-father succession is well attested in tenth-century Wales,
and the kingdoms of Gwynedd and Dyfed were transmitted undivided, though
not without intra- familial conflict. It seems likely that some if not all these fea-
tures resulted from contacts with the English: Hywel Dda was not tire only
Welsh king of this period to spend time at the the court of an Anglo-Saxon
king. But the contacts were very often warlike, as the English raided and plun-
dered and imposed tribute. Poets bemoaned these attacks and invoked heroic
resisters of old. Wales’ emergence as a slightly more united but distinctly more
self-conscious political community was as much a defensive reaction to
English aggression as flattering imitation of English ways.

The simultaneously attractive and repellent power of aggressive and ambi-
tious neighbours can be seen in Scandinavia too. In Norway the practice of
royal law-making is attested for the first time in die tenth century, on the part of
a king who had been brought up at iEthelstan’s court. The OE word bird for
royal retinue was borrowed apparently in the tenth century into Norwegian and
also Danish. Englishmen played a series of important roles in the tenth-
century Christianisation of Norway, and perhaps (then or slightly later)
Denmark and Sweden too. In Denmark, however, the conversion of King
Harald in the 960s was tire result of the freelance efforts of a German priest.
The archbishops of Hamburg-Bremen tried to keep the later Danish bishop-
rics as their suffragans. Official Ottoman influence is hard to find. Rather, it was
Harald who promoted Christianity for his own purposes. He banned pagan
rites; and archaeological evidence shows pagan burial practices abandoned by
tire end of the century. Nevertheless it is not until the twelfth century that
ecclesiastical forms of royal inauguration are documented in Denmark. Harald
modernised selectively. On the grave-monument he erected at Jelling for his
parents he used Christian but also other, indigenous motifs. He chose this
medium, in this dynastic cult-site, for the claim to have won ‘all Denmark’ and
Norway as well. The claims were closely linked. Norway emerged as an inde-
pendent kingdom when the Ottomans temporarily subjected the Danes to
tribute-payment. Thus, just as the Danes had reacted to Frankish, then
Ottoman power, so Norwegians reacted against Danes. Harald’s re-establish-
ment of overlordship of Norway marked a point of conjuncture when both
cycles of contact favoured Denmark: Norway was racked by dynastic disputes,
while in the later 970s Ottoman resources were concentrated on Italy, then in
983 weakened by Slav rebellion and, again, by die dynastic problem of a minor-
ity. Two generations later, the Norwegians reasserted their independence under
their own king. Harald’s power thus depended on external relations. As for
resources within tire kingdom, archaeological evidence makes it certain that

Cambridge Histories Online © Cambridge University Press, 2008

Rulers and government

IIC,

several impressive timber forts (the largest with an internal diameter of 240 m)
were constructed c. 980 to a uniform pattern in Jutland and the eastern islands.
Harald clearly mobilised very large quantities of manpower and timber and, so
the uniform structures suggest, did so directly through his own agents, rather
than relying on local aristocrats as middlemen. The forts’ location inland rather
than on sea-coasts, and their circular form and interior symmetry of design,
suggest that, whatever other (military, economic) functions they had, they were
aimed at die Danish population as symbolic centres of royal power. According
to later medieval sources, Harald’s burdensome demands provoked rebellion.
His style of rulership may also have been resented. When he was ousted and
died in 987, the rebels’ choice of successor was his son Sven -who let the forts
fall into disrepair and set about amassing silver through foreign raiding, espe-
cially in England.

Poland and Hungary also are tendi-century examples of state-formation on
the periphery of empire. Bodr owed their conversion to missionaries from
Germany; and direct Ottoman political input continued sporadically impor-
tant. Boleslav Chrobry in Poland and Vajk-Stephen in Hungary seem to have
promoted Christianity under their own steam, exploiting the ideology and
organisation which could help them impose and maintain some kind of
authority over erstwhile peers, and at the same time resist Ottoman pressure.
On the other hand, they were interested in acquiring a veneer of legitimacy
through imperial approval: hence their welcome for Otto III in 1000— 1. Like
Harald in Denmark, Boleslav and Vajk-Stephen were interested in the sym-
bolic representation of their power. They acquired royal titles and attributes,
and holy lances of their own. Regular tribute-paying to the Ottomans was dis-
continued, though the Poles paid again in the eleventh century (as did the
Bohemians). Instead they became sometimes uneasy allies. Culturally, they
assimilated into Latin Christendom. Though they had never been part of the
Carolingian empire, from now on they displayed key forms of the Carolingian
inheritance: its diplomatic, its rituals of rulership, its capacity for ethnic inven-
tion - and above all its religion. That was the face they turned westwards. On
other frontiers, they copied the west’s aggression, imposing tribute on and
taking slaves from their own northern and eastern neighbours.

The textbook opposition of aristocracy to monarchy would have surprised
tenth-century people. Aristocratic power depended on and imitated the power
of kings, while kings for their part could conceive of no government that did
not involve aristocratic collaboration. Discussions, negotiations through face-
to-face encounters, participation in ritual, the use of honour and shame, the
deployment of personalised wrath and grace: these had all been fundamentals
of Carolingian government. In die old Frankish lands, there is more evidence
for all the above in the tendi century. Writtenness, however, takes a lower

Cambridge Histories Online © Cambridge University Press, 2008

120

JANET L. NELSON

profile. There are no more capitularies. It is hard to say how far such differences
of form entailed differences of governmental substance. If participation in
written culture had been very important in the ninth century, in west Francia
especially, then the decline of written communication perhaps caused a sense
of alienation. But other non-written forms of communication and participa-
tion may well have compensated, and had surely anyway always been impor-
tant. Local courts are poorly documented; but a rare document on political
relations in Poitou, the Conventum of William of Aquitaine and Hugh of
Lusignan, c. 1020, refers to such placita , 66 It is hard to find any royal input at this
level though. The count of Angouleme and his faithful men governed without
reference back to the king. They were involved in direct relations with Duke
William of Aquitaine, however, and he occasionally saw the king: Robert the
Pious was his cousin and in 997 actually made a rare trip into Aquitaine to help
William besiege a local enemy. 67 Richer’s wording suggests, though, that the
king acted to help a kinsman rather than to restore order generally.

Continuities can be seen in the role of the church and of churchmen. Kings’
rapport with the church became more intimate, in some ways more comfort-
able. Monks and bishops were more prominent than ever before in the conduct
of tenth-century diplomacy. As in the Carolingian empire, the church helped
give territorial as well as ideological shape and focus to realms old and new. In
west Francia, Rheims played a key role throughout tire tenth century in sustain-
ing the monarchy and, in 987, in guiding its transfer from one dynasty to
another. In multi-centred east Francia, several churches shared this role, Mainz
and Magdeburg prominent among them. Canterbury, Winchester and
Worcester in England, Gniezno in Poland, Gran in Hungary, and even St
Andrews in Scotland, played analogous roles. But nowhere did monasteries
and convents make such a substantial contribution as they did in east Francia to
sustaining the monarchy through the provision of hospitality and tire exercise
of local jurisdiction through advocates.

Alone among the former Carolingian realms, the kingdom of Italy offers a
different picture. Here ecclesiastical power, locally strong but divided against
itself, provided no cement. In any case, in Lombard, then Carolingian, Italy the
institutions of secular government had been peculiarly strong, kings had legis-
lated vigorously, and that legislation continued to be applied in courts.
According to some modern historians, the state survived well in Italy through
the tenth century. It is true that the palace of Pavia continued to function, and
that presupposed organisation of royal economic resources. In 947, Berengar
II of Italy collected a poll-tax to pay off die Hungarians (diough he allegedly

66 Conventum inter GuillelmumAquitanorum comes etHugonem Chiliarci)um\ ed. Martindale, pp. 545, 547.

67 Richer, iY/r/orehtf iv, 108, p. 330.

Cambridge Histories Online © Cambridge University Press, 2008

Rulers and government 1 2 1

kept most of the proceeds). 68 Nevertheless, this is only part of die story. It is
from the tenth-century Italian kingdom that the clearest evidence comes for
systematic dismantling of structural bonds between centre and localities. The
semantic evolution of die word districtio is revealing here. In early medieval
legal texts, it had meant punishment or legal coercion. In the tenth century it
increasingly often had the more general sense of jurisdiction, such as that exer-
cised by bishops or counts; and the word occurs in this sense in Italian charters
in which kings granted away such powers. In 940, Kings Hugh and Lothar
granted to a count an estate ‘with all districtio and all public function and capac-
ity to hear legal suits which previously our public missus was accustomed to
perform’. 69 The unique longevity of the Italian state, then, may have been
exaggerated. Here, as elsewhere, what can be seen is devolution protracted to
local autonomy. At the same time, the word districtus, very rare before the tenth
century, and originally meaning ‘coercive action’, acquired in Italy die sense of
‘an area of jurisdiction, especially considered as a source of revenue’ - so, a
governmental ‘district’.

If tenth-century counts can hardly be considered any longer as ex officio
agents of royal government in Italy, the same is true elsewhere in the old
Carolingian kingdoms. The title count, generally hereditary even in die
Carolingian heyday, was no less so in the tenth century. Nevertheless, an active
Ottoman king gave his formal approval when a count treated his offices ‘like an
inheritance’ and divided them among his sons. 70 Even in west Francia, before
such a division occurred on the death of die powerful count Heribert of
Vermandois, the count’s sons went to the king and were ‘benignly received’. 71
In neither east nor west Francia could kings readily remove uncooperative
counts, though if a count actually rebelled he might be attacked by a king on
grounds of unfaithfulness, his lands ravaged, and eventual capitulation
rewarded with nominal reinstatement. Most counts were simply too far away to
be involved in regular contact with kings at all. In east Francia there was,
however, a higher level of potentate: dukes and margraves. Though these too
were hereditary, the king formally appointed them, and may also have had
some say in their marriages. In some cases, these men became relatives by mar-
riage of the ruling dynasty. Timely interventions over die marriages of ducal
widows and daughters were important means of extending royal influence into
duchies, as Otto I showed in Bavaria in 947 when he arranged the marriage of
his brother Henry with the daughter of the defunct Duke Arnulf. The facts of
demography - die failure of male descent and the frequency of claims through

68 Iludprand, Xlntapodosis v, 33, p. 1 51- 69 D. Hugh 53, pp. 160— 1.

70 Adalbert, Reginonis Continuation s.a. 949: ‘ U to comes obiit, qui permissu regis, quicquid beneficii aut

prefecturarum habuit, quasi hereditatem inter fiiios divisit . . .’

71 Richer, Historiae 11, 37, p. 186.

Cambridge Histories Online © Cambridge University Press, 2008

122

JANET L. NELSON

women — offered die king occasional opportunities to intervene in the trans-
mission of property, and sometimes to claim reversion to the royal estates.

It goes without saying that though die western economy was patchily mon-
etised, no king was in a position to pay salaries. Government was of a type that
could be largely carried on without die activity of full-time specialists.
Chanceries, for instance, consisted of a handful of clerics working as royal
notaries quite intermittently. Government worked in the main through face-to-
face contacts and direct consultations between powerful people. Yet this did
not mean that kings lacked any agents at all. They could deploy the personnel
of their own households, their ‘men’ {homines, domestic i) and their ‘servants’
{ministri, ministerialed). Even in west Francia, at least in a restricted area around
Rheims and Laon, kings could command the services of castellans. 72 In 926, a
tax was raised ‘publicly throughout Francia’, presumably collected by the king’s
own men, to buy off the Northmen: a feat of organisation which, even if we
scale down Flodoard’s ‘Francia’, implies the services of a number of royal
agents. 73 Many senior churchmen too, in however partial and part-time a way,
worked on behalf of kings. When kings granted and confirmed immunities
(and, significantly, such confirmations were often sought) to bishops and
abbots, they had no intention of signing away absolutely the judicial rights and
profits involved. The very closeness of their personal relationships with these
churchmen — over whose appointments they had generally exerted some influ-
ence - meant that kings, by and large, could successfully demand from
churches tire two things they required: hospitality and the service of warriors
maintained on church lands. In supervising such services, churchmen indeed
worked for the king. And they sometimes worked as a team: in east Francia
especially, the continuing (if intermittent) Carolingian practice of holding
large-scale synods under royal auspices kept in being a sense of a regnal
church. West Frankish bishops preserved something of the same esprit de corps',
and some of them manifested it, eventually, in peace councils. In Italy, despite
the pleas of individual bishops like Rather of Verona, such collaborative
efforts were lacking. Rather was a Frank from Lotharingia, overly dependent on
personal ties with the Ottoman court.

There are traces of royal agents in some outer-zone kingdoms as well.
Counts, and local judges called saiones, functioned regularly as court-holders in
Leon-Castile. Sometimes, however, saiones are explicitly described as ‘of the
count’. The kings’ power was based in Leon. They ruled Galicia and Castile via
personal dealings with the aristocracy of those regions; and the clearest sign of
royal authority there is the presence of regional aristocrats at the king’s court.

72 Richer, Historiaen, 7, pp. 136/8.

73 Flodoard, Annates, s.a. 926, p. 34: ‘exactio . . . publice fit per Franciam’.

Cambridge Flistories Online © Cambridge University Press, 2008

Rulers and government

123

In Denmark, by the close of the tenth century, adminstrative districts (syssel)
had apparendy been established in Judand, the core-area of Harald’s kingdom:
probably an important shift away from die kind of personalised, non-territor-
ial organisation which survived in Iceland into the twelfth and thirteendi cen-
turies. Harald was able to get impressive fortifications built on at least four
sites, perhaps by imposing forced labour. Military followers may well have
functioned as all-purpose agents. In Wales the prologue to Hywel Dda’s legal
collection asserts that it was put together by skilled lawyers and clergy acting on
the king’s orders. 74 But that same legal evidence suggests that die apprehending
of criminals was landowners’ responsibility. Of special royal agents there is no
trace. Kings with their retinues could impose demands, and sanctions, directly
in kingdoms barely 9 5 kilometres across. Interestingly, the only clear evidence
for royal law-enforcers and tax-collectors in the Celtic world of this period
comes from Scotland, which covered an area seven times larger than the little
kingdoms of Wales.

One key activity of government may be identified as the mobilisation and
application of force in conflict-management. It was difficult even for English
kings to apply force effectively as a sanction against influential uncooperative
subjects. fEthelred, as already noted, had to summon two ‘great meetings’, and
took years, before he was able to make Wulfbald, and later his widow, obey the
royal command. Wulfbald’s story has been taken to show ‘the extraordinary
feebleness of the government he defied for so long’. 75 It could be argued,
instead, that by tenth-century standards, the outcome of this case showed
extraordinary royal effectiveness. When fEthelred finally got his way, though, it
was not by overwhelming force. True, he was able to use exile successfully
against a string of magnates. But in all these instances, he could operate only
with substantial aristocratic support. This he relied on when confronting exter-
nal enemies too, even if local peasants might rally to defend their homes. At
Maldon in 991, the army defeated by the Vikings consisted essentially of the
followings of Ealdorman Byrhtnoth and his kin and local allies. Contemporary
west Frankish kings were generally able to act against faithless castellans in
their own neighbourhood, by the standard military tactics of ravaging and
besieging. A king, leading a military retinue probably numbered in tens rather
than hundreds, was militarily effective therefore within a limited area and with
limited objectives: in fact more or less on a par with tire principes and even castel-
lans who were his routine opponents. West Frankish kings seldom had to con-
front external enemies, hence did not - could not - invoke ‘the defence of the
realm’. The Ottomans faced frequent external threats; and organised military

74 The Lem 1 of Hywel Dda, p. i : ‘six men from every cantred in Wales’.

75 Whitelock in EHD, p. 47; cf. Keynes (1 991), pp. 78—9.

Cambridge Histories Online © Cambridge University Press, 2008

124

JANET L. NELSON

resources accordingly. They too relied on their own military followings, which
may have consisted of up to a thousand men, 76 but in addition asked their prin-
cipes, ecclesiastical as well as lay, for contingents in particular campaigns: thus,
as at the Lech in 955, an east Frankish army might exceptionally have been
brought up to a strength of several thousand rather than a few hundred.
Miesco of Poland’s remarkable success depended not least on his large follow-
ing - of 3000 men, according to an Arab source 77 (almost certainly exaggerat-
ing - Arab historians were as liable as westerners to be fazed by numbers). The
Ottomans’ military dependence on aristocratic and ecclesiastical contingents
was no innovation, for Carolingian armies had arguably been similarly
recruited except in cases of local resistance to outside attack. It may be signifi-
cant though that tenth-century writers labelled these contingents by their
leaders as well as (as in the past) by their local origin. The military power of
magnates and retinues loomed large in the Ottomans’ kingdom as elsewhere.
The social power of these men, after all, rested on their status as a warrior elite.
Kings’ conduct of war thus depended on consent, and tire mobilising of faith-
ful ones and friends. Karl Leyser has suggested that warfare might be seen as a
normal state, and policy its pursuit by means that were only incipiently
violent. 78

As in the Carolingian period, one thing above all held political systems
together, and kept peace of a kind: that was tire assembly - to be seen as a
regular and sanctioned event, a social conjuncture, a set of interactions, the
centre of a field of centripetal force, in short, as an institution. Assemblies
were often located at cult-centres. Regnal communities were formed and rein-
forced through meetings, and by tire forging and reforging of personal links,
between the king and his great men. These were occasions for multiple forms
of collaboration and bonding. The banquet was a cross between working lunch
and club dinner, with strong undertones of the public house. For tenth-
century rulers it was not just useful, but essential, to combine political discus-
sions with social exchanges: colloquia, serious conversations, indeed consisted
of both. Forum for personal meetings and the forming of personal relation-
ships through the exchange of gifts and services; marriage market; a job
market of sorts; ritual theatre; exercise-ground for companionship in war: the
assembly was all these - and all these were more important than what we call
bureaucracy in maintaining the community of the realm.

One striking example is the assembly at Senlis in 987, where Hugh Capet’s
succession to the west Frankish kingdom was settled ‘after die opinions of
different participants had been collected’. Richer’s account telescopes various

76 The estimate of 1000 is that of Werner (1979), chapter 111, p. 828.

77 Ibnjaqub, cited by Heather (1994), p. 62. 78 Leyser (1994), ch 3.

Cambridge Histories Online © Cambridge University Press, 2008

Rulers and government 1 2 5

arguments into die archbishop’s speech: 79 positively, Hugh’s physical qualities
(nobility of body) and personal qualities (wisdom, credibility, big-heartedness)
were required; negatively, Charles of Lorraine excluded himself by his service
to an external king (Otto II) and his marriage to a woman of merely knightly
status. Absent from the account of the archbishop’s speech, but observed
shortly afterwards by Richer himself, was the crucial consideration against
Charles: his need, as king, to reward and endow his hitherto footloose follow-
ers - and dtis could only be done at the expense of sitting tenants. The ‘haves’
considered dieir interests, and voted for Hugh, against the Carolingian candi-
date.

The linked assemblies of 973 in Bath (attended by ‘a great company 5 ) and
Chester offer another, English, example. 80 These occasions were imposing
scenes of royal ritual, but at the same time forums for hard-headed political
discussion about coinage reform, law, relations with neighbours and tributar-
ies, contacts with the Ottoman court. Further examples frame the reign of
Otto I: in 938, ‘a decree went out from the king that an assembly of the whole
people’ (Widukind’s phrasing here echoes St Luke) should meet at Steele near
Essen to consider a difference between the customary inheritance laws of
Franks and Saxons. The agreed outcome reasserted ‘royal power’ as well as
being in the interests of ‘associates’. 81 In 973, Otto held what was to be tire last
Easter assembly of his reign at Quedlinburg. ‘A multitude of diverse peoples’
attended to celebrate Otto’s return from Italy: at Merseburg for Pentecost, the
peoples in attendance included those beyond the eastern frontier, and also
‘envoys from Africa’. 82 This last touch evoked Carolingian glories, but it was no
fantasy. In contemporary accounts of dealings with peoples beyond the fron-
tiers are hints of Latin Christendom’s new self-consciousness.

‘ourselves as others see us’?

Cultural contacts often highlight difference. When die nordr Italian Liudprand
visited Constantinople at Eastertime 9 50 as envoy for King Hugh, he was enor-
mously impressed by the annual distribution of money to the holders of offi-
cial tides, rank by rank: diose of the higher echelons ‘needed assistance to drag
their money laboriously away’. 83 This was only possible in a regime that taxed
on an extensive scale, that is, one that had highly developed mechanisms for
creaming oft' and redistributing wealth. By Liudprand’s next visit, on behalf of
Otto I in 968, his admiration had gone sour. His famous description of

79 Richer, Historiae iv, 1 1 , pp. 158/62.

80 ASC, versions ‘d’, ‘e’, pp. 227— 8;^lfric, Life of Srnthin , trans. EHD , pp. 927—8.

81 Widukind, Res gestae Saxonicae 11, 10, p. 62. 82 Leyser (1994), pp. 96—7.

83 Liudprand, Antapodosis vi, 10, pp. 157—8.

Cambridge Histories Online © Cambridge University Press, 2008

JANET L. NELSON

Il6

Nikephoros Phokas’s court ceremonial is not objective reportage: Liudprand
parodies the majestic and mystifying self-representation of the basileus , 84 Even
if you did manage to come into his presence, there was little of the colloquium
here. Prostration was not a posture conducive to serious conversations. Basil II
was allegedly advised not to be approachable. 85 Liudprand’s eventual scorn for
the style of Byzantine monarchy conveyed more than a change of personal
attitude: the subtext was a new western assertiveness, even arrogance, towards
‘the Greeks’ and their pretensions to romanity.

Like Constantinople, Cordoba in the tenth century was a very big city: both
were centres of extensive territories with clearly defined frontiers and cross-
ing-points patrolled by customs-men (as Liudprand learned to his cost in 968).
The two regimes commanded standing armies of many thousands, paid for
out of tax revenues. In Basil II’s reign, the army contained a central component
of professional foreign troops known as ‘Varangians’ (cf. ‘Franks’). The mili-
tary strength of al-Andalus depended on foreign slave-soldiers (many of Slav
origin, reflected in the Arabic word for slave), highly trained from youth and
highly efficient. Such forces could operate far afield: an army was sent nearly
700 km from Cordoba to sack Compostella in 997; Nikephoros Phokas had
dispatched a fleet 900 km from Constantinople to capture Crete in 961. Like
die basileus , the caliph ran a vast court, geographically distinct from the capital
city itself. ‘Abd al-Rahman III had only recently installed himself in die newly
built palace at Madina al-Zahra 5 kilometres outside Cordoba, when he sent
envoys to Otto I to initiate an exchange. Both caliph and king were concerned
to suppress piracy in the western Mediterranean and along the Provencal coast.
The result, in 953, was a return embassy led by Abbot John of Gorze in
Lotharingia. John’s Life was written up some two decades after his death in 974
by the abbot of Saint-Arnulf, Metz, who had known him and used the aged
abbot’s reminiscences. The Life was dedicated to the archbishop of Trier, and
its expected audience consisted of Lodiaringian churchmen. Here, in however
stylised a form, is purveyed an Ottoman view, or more precisely, a Lotharingian
view, of Cordovan power. 86

John was amazed at the sheer scale of Cordovan officialdom, the amount
of paperwork required for every kind of communication, even the size of
documents. A thicket of bureaucracy seemed to block access to the caliph
himself. John waited for the best part of three years in an official hostel at
some distance from the palace before a meeting with the caliph was arranged.
The vast scale of the palace complex — impressively confirmed by recent

84 Liudprand, Relatio. 85 Michael Psellos, Chronographia, trans. Sewter, p. 43.

86 John of Saint-Arnulf, Vita Iohannis abbatis Gor^iensis, cc. 118— 36, pp. 371— 7; partial English translation
Fletcher (1992), pp. 67-8.

Cambridge Histories Online © Cambridge University Press, 2008

Rulers and government

I2 7

excavations — evidently made a profound impression on John. Slave soldiers,
infantry and cavalry, guarded access to it, and performed imposing military
exercises, making their horses rear up ‘to put fear into our people’. John
approached through outer courtyards ‘covered with the most costly rugs and
carpetings’, to encounter the caliph ‘on a dais, alone, almost like a god access-
ible to none or very few, and sitting, not on a throne or seat as other peoples
do, but on a cushion’. His offer of the inside of his hand to kiss was a tremen-
dous honour ‘not customarily granted to any of his own people or to foreign-
ers’.

What purports to be a Cordovan view of the Ottomans is also recorded by
John’s hagiographer. Whenjohn asserted that his lord, King Otto, must be the
most powerful ruler in the world in terms of men and territory, the caliph
riposted:

your [king] does not keep for himself alone the power of his strength \potestas virtutis
suae] but rather he allows each of his men to wield his own power: he shares out the
regions of his realm amongst them, thinking thus to make them more faithful and more
subject to him. But the outcome is very far from that! What is nurtured is pride and
rebellion . . . and the rebels call the Hungarians into the midst of their regna to lay them
waste . . .

At this point, unfortunately, the Life, in the single manuscript, breaks off before
giving John’s rejoinder. Presumably after condemning magnate - and not least
Lotharingian — disloyalty (remember for whom the Life was written) through
the mouth of the caliph, the hagiographer ended on a high note. By the later
970s, the Hungarian threat was over; and the victory of the Lech, to which
Lotharingians contributed, had triumphantly affirmed Ottoman success in
955. The caliph’s critique could thus be seen to reflect false generalisation from
the revolt of 954, and thence a misunderstanding of Ottoman government.
The hagiographer invited his audience not to envy caliphal authority, not to dis-
parage Otto’s regime, but instead to relish that extensive devolution of power
which was the most striking feature of kingship in tenth-century Latin
Christendom. Here was a Lotharingian view refracted. The Life of John of Gor^e
was written, like Montesquieu’s Persian Letters, not as true ethnography but as
commentary on home politics, and for home consumption. Its purpose was to
celebrate not onlyjohn’s holiness, exemplified in his stout refusal to dress up
for his audience with die caliph (who allegedly admired his ‘unyielding strength
of mind’ and declared his willingness ‘to receive him even if he comes dressed
in a sack’), but also the home regime itself. Power-sharing, and a measure of
internal conflict, could coexist with a powerful regime - one which could boast
the services of a strong-minded saint.

Earlier on in die story, the hagiographer explained why the caliph initially
refused to receive John’s embassy: the letters sent by Otto contained words that

Cambridge Histories Online © Cambridge University Press, 2008

128

JANET L. NELSON

blasphemed the Prophet. This news got out, and the Cordovan populace
became angry, reminding the caliph that if he did not punish blasphemy with
death, by Muslim law he himself deserved to die. It was a tricky situation. While
John was kept waiting, die caliph sent a Spanish-Christian bishop, Reccemund,
to Otto to obtain new and inoffensive letters. The author of die Life affects to
register John’s surprise at the caliph’s inability to modify the law, in this case to
exempt from a punishment, on his own authority. That was precisely what
King Otto not only could do but was expected and invoked to do by his power-
ful subjects. An essential trait of east Frankish rulership, in other words, was
the capacity to judge, and to practise equity - to exercise discretion in applying
the law. A ruler with such authority might lack the huge bureaucracy and parad-
ing soldiery of Madina al-Zarha. In direct contact with his magnates in the rel-
atively intimate atmosphere of his hall, Otto had no need of the mediation of
hosts of officials like those who transmitted successive explanations for delay
during the three years John was kept waiting in the environs of Cordoba, no
need of serried ranks of slave-soldiers. Nevertheless Otto was a mighty ruler:
the hagiographer invites readers to admire something very different from cali-
phal rulership.

Last but not least, the hagiographer implicitly denounces the time-servers
who took the orders of tire infidel caliph: the Jew Hasday ibn Shaprut, and the
Mozarab Christian Reccemund. These men lived under and cooperated with a
Muslim regime: ‘provided no harm is done to our religion, we obey them in all
else and do their commands’. John stood for an alternative. He was defiant, an
outspoken witness for Christianity, whom only the threat of causing a blood-
bath among Spanish Christians deterred from martyrdom. Like Liudprand of
Cremona, the author of John’s Life breathed the spirit of a Latin Christendom
reformed and militant, and - for good or ill - of a monarchy to match.
Gregorianism did not have all the best tunes.

The caliph’s response to the tenth-century Ottoman kingdom prefigured
that of many modern historians. What kind of a kingship is this which permits
so many sharers? How could this be dignified by the name of government?
Political anthropologists, familiar with the traditional states of Africa, might
take a more percipient, and a less anachronistic, view. For a relevant early med-
ieval comparison and contrast, we might look also at the Icelandic pattern of
decentralised power and shared attitudes about lawfulness, including the legiti-
macy of rebellion and private justice, in a kingless system. This chapter has
explored systems which in many ways resembled Iceland, but which func-
tioned with and through kings and kingship. Hence these were states of a dis-
tinctive kind, one unappreciated by the hagiographer’s imagined caliph.
Constantinople and Cordoba alike offer instructive comparisons with the west,
but they were fundamentally remote from it. Their writers have left no records

Cambridge Histories Online © Cambridge University Press, 2008

Rulers and government

129

of their reactions to the Christian realms, and so we cannot get acquainted with
the world of tenth-century kingship and government through foreign eyes. In
the end, our most useful go-betweens, for all their limitations, are the contem-
porary texts - and especially die historians - of die Christian realms them-
selves. It was, after all, their world.

Cambridge Histories Online © Cambridge University Press, 2008

CHAPTER 5

THE CHURCH

Rosamond McKitterick

the history of the church in Europe in the tenth and early eleventh centu-
ries is essentially the history of many local churches, in which the dominant
role in secular ecclesiastical and religious life was played by die bishops. Only
occasionally can large-scale collective activity be observed; for the most part
the very different challenges to religion from within and without the Christian
world and the responses to them on the part of various members of the eccle-
siastical hierarchy in western Europe mean that it is above all local preoccupa-
tions and regional differences that are reflected in the surviving evidence.
Nevertheless, die evidence taken as a whole, that is, the synodal legislation and
canon law collections, lives of bishops, histories of sees and monasteries,
liturgy, music, accounts of saints’ cults, books containing patristic and
Carolingian theology and biblical exegesis produced for use within ecclesiasti-
cal institutions, theological treatises, polemical pieces d’occasions and incidental
references in the narrative histories of the period, reveals not only lines of con-
tinuity with the ninth-century Carolingian church but also many elements of
coherence and unity within the remarkable diversity of die tenth-century
church. Not all of this coherence can be attributed to die links with monasti-
cism and monks throughout western Europe on the part of the secular clergy,
though the zeal for monastic reform undoubtedly was a common bond right
across Europe. The evidence itself, moreover, presents particular problems of
interpretation, for much of it was designed to present ideals and norms,
whether of saints’ and bishops’ behaviour, of what was expected of the laity in
the Christian observance, or of the interaction between the clergy and the laity,
which do not necessarily provide a faithful reflection of reality.

Throughout this chapter, therefore, as well as determining the principal
activities and achievements of the tenth- and early eleventh-century church,
the relationship between the extant sources and the reality they purport to
describe will be assessed in relation to the pastoral and political role of individ-
ual bishops within their dioceses, the collective activity of the clergy in councils

1 30

Cambridge Histories Online © Cambridge University Press, 2008

The Church

1 3 1

and synods, die function of die law and liturgy of die church, and manifesta-
tions of lay piety.

The organisation of the church in the tenth and early eleventh centuries main-
tained the structures established in the early years of die Christian church and
particularly that of the Frankish church in the preceding three centuries, whose
principles had originally been based on Roman imperial administrative units.
Ecclesiastical provinces, headed by die metropolitan or archbishop, comprised
a number of dioceses, whose bishops were under the jurisdiction of their met-
ropolitan. Thus, for example, the province of Trier included the dioceses of
Toul, Metz and Verdun. Each bishop in his turn had the charge and care of the
clergy and laity of his diocese entrusted to him. Although in principle elected, a
bishop was often in practice nominated, either by the existing incumbent, or by
lay authorities, the first of which methods was uncanonical and the second, on
occasion, ill judged. Kings and other rulers interfered to varying degrees in
episcopal elections, and such interference could work to the church’s advan-
tage or disadvantage depending on the wisdom of the choice and the degree to
which a position could be abused in terms of the self-aggrandisement of the
bishop or members of his family. The political complexities such interference
could create are discussed below in relation both to the papacy and to particu-
lar sees, such as Rheims. Certainly the interest taken in episcopal appointments
on the part of local magnates tended to preserve their position within the ranks
of the social elite. Robert tire Pious, on the other hand, according to Radulf
Glaber, ‘took great care to fill [any bishopric in his realm which lost its incum-
bent] preferring a suitable pastor of low birth to a nobleman steeped in the
vanities of this world’. 1 It is striking how many members of die episcopate as a
whole, even if we discount the hyperbolic attribution of ‘noble’ origins to
many bishops by their biographers, were closely related to die secular rulers of
the regions they dominated in ecclesiastical matters. Leadership in secular and
spiritual matters in many areas, such as Rheims or Metz, as will be seen below,
was a thoroughgoing family concern.

There has been much criticism of this situation, both on the part of contem-
poraries and by modern historians, for if abused, power and, crucially, prop-
erty could get into the wrong hands. There are many instances, such as in the
dioceses of Metz under Bishop Adalbero II, where church lands had been
given out by his predecessors, or of Laon under Bishop Rorico, where episco-
pal estates were held by the bishop’s kindred. The treatise Dialogus de statu
sanctae ecclesiae of Malcalanus ( c . 962) refers to the bishop’s obligations to look
after the interests and immediate necessities of his kinsmen and the poor, and

1 Radulf Glaber, / listoriae nr, 7.

Cambridge Histories Online © Cambridge University Press, 2008

1 34

ROSAMOND McKITTERICK

the difficulty of determining what was immoveable and inalienable, especially
in relation to episcopal properties already apportioned. Fichtenau has warned
against too rigid an interpretation of the terms of such gifts, for they may not
always have entailed a permanent loss to the church. Effective lay control of
ecclesiastical property gave die lay owner considerable power within a diocese,
but if the bishop were of sufficient standing he could hold his own.

There is no doubt that the church could provide both an element of stability
within a polity, and an excellent supply of able and educated personnel who
could assist in the process of government and administration. In the Saxon
kingdom in particular, such work entailed service in the palace chapel and the
royal writing office for the production of charters and laws. Those who had
worked at court, moreover, subsequently became bishops or abbots elsewhere
in the kingdom. But die position of the Frankish and Saxon rulers in relation to
the clerical hierarchy had its origin in die careful equilibrium more or less main-
tained between Germanic rulers in nordiern Gaul since the conversion of
Clovis and enhanced and augmented under the Carolingians in the eighth and
ninth centuries. Much of the character of the tenth-century episcopate in rela-
tion to secular rulers, notably in the Frankish heardands, Wessex and perhaps
Leon, can be accounted for, moreover, not only by the extraordinary combina-
tion of political expansion and instability recounted in the various chapters in
this volume, but also by the various ways in which emergent new states sought
to define and express their identity both in their own terms and in relation to
the older, primarily Carolingian, institutional framework and norms to which
they were heirs. Thus it is unlikely that quite such a clearly defined ‘system’ or
consistent identification of die church as a counter-balance to the aristocracy
existed within the Ottoman kingdom in particular, let alone elsewhere, as was
once imagined. Such a concept underplays the interlocking and interdependent
nature of interest groups and institutions, both lay and ecclesiastical, as well as
die customary modes of procedure and social behaviour throughout early
medieval society. Any fundamental antagonism between the nobility and the
church anywhere in western Europe is hardly to be credited in relation to
die available evidence. The German evidence is of a piece in its refiection of
die determined use bishops and rulers made of each other and may well reflect
a far more coherent policy than is evidenced elsewhere. It was still largely from
die ranks of die nobility, however, that bishops were recruited. It remained the
continuing wealdi and patronage of the nobility on which the survival of the
church depended.

Reform of the church in diis context, whether advocated by bishops, abbots
or princes, was essentially about control and the degree to which bishops or
laymen could be permitted to exert the influence and power that each wished.
In asserting independence from lay interference, and in defining wholly eccle-

Cambridge Histories Online © Cambridge University Press, 2008

The Church

1 3 5

siastical norms for ecclesiastical institutions, tire church was arguably, there-
fore, in danger of cutting off its nose to spite its face in the interest of short-
term imagined advantages. In wishing to assert too heavy-handed an interest,
however, secular rulers were equally in danger of upsetting a delicate equilib-
rium between secular and spiritual authorities and their representatives, and
this is nowhere clearer than in tire history of the popes in the tenth century out-
lined below. It is, furthermore, not possible to describe the history of die
tendi-century church purely in terms of a loss of such equilibrium, or even
more creditably, a fight to maintain it. There was throughout the kingdoms of
western Europe an ever-changing balance and uneasy shifting of local and
regional interests, like loose ballast in die hold of a ship.

Whatever the source of his patronage and impetus to gaining office may
have been, the bishop’s spiritual office was confirmed by the church in that he
was consecrated by his fellow bishops and normally selected by the chapter.
The bishop presided over the cadiedral familia, whose members had pastoral as
well as liturgical obligations within the episcopal city. The bishop’s prerogatives
as well as his sphere of jurisdiction and pastoral obligations were increasingly
emphatically defined in die course of the tenth century. The synod of
Hohenaltheim (916), for example, devoted a number of paragraphs not only to
the definition of a Christian bishop (with reference to both Old and New
Testament statements on the function of a priest) but also to his duties and
expected behaviour and his role in safeguarding the privileges of die church. 2
In his Praeloquia, moreover, the curmudgeonly Rather, bishop of Verona, twice
deposed and restored to his see because of, in his opinion, political opposition
and persecution, is eloquent on the respective prerogatives of the bishop and
secular rulers. A bishop is a powerful and active man. He should serve
emperor, archbishop, clergy and laity and travel diroughout his diocese, con-
ferring widi the most important clerics and laity about what was to be done in
order to do justice to everyone. Radier invokes a veritable phalanx of Old
Testament priests and prophets to support his ideal of the exemplary bishop,
whose virtues remain vaguely defined even if his possible array of vices is terri-
fyingly detailed.

From the end of the ninth century, it is likely that there were also officials
such as the archdeacons to assist the bishop’s work in die city, and deans who
would have particular areas in the countryside under their charge. In addition
there was the parish system, increasingly indicated in such sources as royal and
synodal legislation, episcopal statutes, charters, narrative accounts and refer-
ences to the collection of tithes (instrumental, for obvious reasons, in the defi-
nition of the territorial boundaries of a parish) from the eighdi century

2 MGH Concilia aevi Saxonici etSalici, no. i,pp. 19—40.

Cambridge Histories Online © Cambridge University Press, 2008

136 ROSAMOND McKITTERICK

onwards. Nevertheless, the parish system was by no means fully or coherendy
organised even in the early eleventh century. Considerable regional variations
also existed throughout western Europe. Interpretation is rendered more diffi-
cult by the fact that the Latin terms for parish and diocese appear to have been
interchangeable. The Aachen capitulary of 8 18-19, 3 preserved in many late
ninth- and tenth-century collections of ecclesiastical legislation as well as in the
widely disseminated capitulary collection of Ansegis, was of particular impor-
tance in specifying parish provision and stipulating die support for the priest.
Some commentary was also provided by Archbishop Hincmar of Rheims
(845-82) in his treatise De ecclesiis et capellis, who adds the information that a
bishop was to be supported in his city work by the archdeacons, and by the
deans in the countryside. Similarly die episcopal capitularies, widely dissemi-
nated on the continent, and die synodal address Fratres presbyteri provided defi-
nitions of the expected work of die clergy.

Parish priests, ordained by the bishop, were directly subordinate and answer-
able to the bishop or his representative on all ecclesiastical matters, whether to
do with discipline, administration of the sacraments, pastoral care, income or
upkeep of the parish churches. Parish churches are associated not only with
smaller units of a city, such as those created by Bishop Burchard in Worms in
the early eleventh century, and with areas in the countryside, but also with indi-
vidual lords’ estates, the so-called Eigenfdrchen. The development of the parish
appears to have been piecemeal, but with many similarities in all parts of
Europe. In Anglo-Saxon England, for example, pastoral care emanating from
the minsters (an Old English word used to refer to both monasteries and parish
churches) was only gradually supplemented by a parochial system consisting of
small rural and urban parishes with a church, a priest, and sufficient endow-
ment to support both. Evidence for its density, particularly in eastern England,
is late and indirect, for it is embodied in Domesday Book of 1086, as is infor-
mation about the number of churches in such towns as Norwich, which had, at
the end of the eleventh century, twenty-five churches and forty-three chapels,
or York, which appears to have had at least fourteen parish churches.

Some notion of the expected duties of a parish priest can be deduced from
such documents as the early tenth-century charters of ordination from Lucca.
These specify that the priest is to celebrate Mass and other offices, maintain the
lights of the church (that is, keep up the supply of candles and the oil for
lamps), obey the bishop and refrain at all times from alienating any church
property. Episcopal statutes, such as those of Archbishop Ruotger of Trier
(915-31) addressed to all the priests of the churches within his care, and
synodal legislation of die tenth century, such as tire synods of Koblenz 922 or

3 MGH Capitularia regum Francorunn, 275—80, no. 138.

Cambridge Histories Online © Cambridge University Press, 2008

The Church

1 37

Trier 927, not only stressed episcopal authority. They also maintained that the
priests were to set an example by their conduct, to administer the sacraments
punctiliously, not to accept payment for baptism or burial, to teach the people
about dieir faith and to exhort them to proper Christian conduct as defined by
the synod. 4

In all these contemporary discussions and decisions concerning the admin-
istration and organisation of the church, little reference is made to the pope
and his role vis-a-vis die churches, provinces and diocesan sees of Christendom.
Archbishops customarily sought the pallium (a band of white wool embroi-
dered with crosses and serving as a badge of office given by die pope to a
metropolitan) from the pope. Some are known to have done so in person, such
as Dunstan of Canterbury and Oskytel of York who received pallia from Pope
John XIII in 960; others were granted it. In the latter case it was presumably
received from papal messengers, just as Archbishop Adaldag of Hamburg-
Bremen received the pallium sent by Pope Leo VII in about 937. Yet it is only
occasionally between about 900 and about 1050 that there is any sign of the
pope asserting universal leadership or even being acknowledged to have
supreme spiritual audiority. A synod of French bishops convened at Saint-
Basle, Verzy, had deposed Arnulf of Rheims and put Gerbert of Aurillac,
master of die school at Rheims in his place on the instigation of Hugh Capet.
At Pope John XV’s attempt to intervene, the French maintained their right to
independence of action, though John XV did succeed in getting Gerbert
suspended at the synod of Mouzon in 995 and his successor Gregory V
restored Arnulf to his see. The popes, in short, played a far from consistent
role. Their actions and policies depended to an extreme extent on their abilities
and on those of the secular rulers with whom they collaborated or on whose
protection they depended. They rarely intervened other than when asked to do
so.

To some degree it appears to be die case that whatever the character of the
incumbent, the standing of the papacy as an institution, as it had been estab-
lished in the course of die eighth and ninth centuries, survived. One major
contributor to this was the papal notariate, which, like the modern British civil
service, or, for that matter, the papal bureaucracy during the fifteenth-century
Renaissance, provided essential stability and continuity in both its production
of documents and its personnel. The notariate was responsible for the issuing
of papal letters and charters on a regular basis, often on request from suppli-
cants. The erratic survival of documents emanating from the papal scrinium or
writing office in the course of the tendi and early eleventh centuries means that
conclusions can only tentatively be drawn. Nevertheless, it would appear from

4 MGH Concilia aevi Saxonici et Salici, no. 4 , pp. 68 — 74 .

Cambridge Histories Online © Cambridge University Press, 2008

138 ROSAMOND McKITTERICK

the names recorded of the scribes producing letters and charters for each pope
that many individual notaries did indeed serve a number of different popes
consecutively, patiently coming in to work in the papal writing office regardless
of the upheavals connected with the papal throne itself. The papal office was
staffed with a number of notaries at any one time, though some predominate in
the documents at particular periods. Stephen and Leo for example were the
scribes of most of the documents produced between 95 5 and 973; it is con-
ceivable that it is the same Stephen who then appears as scribe of papal char-
ters right through to 992. Under Gregory V and John XVIII the principal
scribe appears to have been Peter, while under John XIV through to Benedict
VIII, Benedict the notary is also very active. Other scribes, such as Nicolas,
Samuel, Gregory, Sergius, Leo, Melchisadech, Antony, John, Adrian, Theodore
and Azzo in the first half of the tenth century and George, Anacletus,
Theodore, Bonizo, Gregory, John or Antonius in the early years of the
eleventh century, make fleeting appearances. The charters witness to contact
being maintained between the papacy and churches in England, France, Spain
and Germany as well as in Italy and Rome itself. No doubt as an accident of
survival, most of the extant charters, in fact, have to do with matters outside
Rome. They witness to the constant stream of requests for papal protection
from such monasteries as San Vincenzo al Volturno, Saint-Martin at Poitiers,
Cluny, Brogne and Quedlinburg, confirmation of ecclesiastical privileges,
especially the designation of metropolitan status and tire conferring of a
pallium, and confirmations of tire position of particular bishops. It is espe-
cially notable how many religious houses sought direct privileges from the
pope rather than from their local rulers. Occasionally, in addition to the
number of charters which relate to political involvement, the pope attempted
to take a moral stand. The lay abbot Hugh, dux of the Franks and lay abbot of
Tours, for example, was abjured not to tolerate the presence of women in the
monastery. A quarter of the 630 papal charters surviving from the period
896-1049 are forgeries. Those of near contemporary date, however, attest to
the degree of habitual reference to the papacy and its authority on a steadily
increasing scale throughout the tenth and first half of the eleventh centuries.

From die perspective of the end of the eleventh century, the tenth-century
papacy nevertheless looks a sorry tale of disgrace, political corruption and
excess, far worse than mere incompetence. Later historians were to term its
period of domination at the beginning of the tenth century by the Theophylact
family - Theophylact, his wife Marozia and dieir sons - as a ‘pornocracy’. This
is to go too far. The most striking feature of die popes in diis period is the
degree to which immediate circumstances were responded to, and how much
attempts to return to past policies or definitions of papal status, notably vis-d-vis
die Frankish or German emperors, had the effect, time and time again, of

Cambridge Histories Online © Cambridge University Press, 2008

The Church 139

hauling the papacy back onto the rails. Table 1 (p. 692) shows the bewildering
succession of popes (striking in its similarities with the succession of fifth-
century emperors), with no less than forty- five in 160 years, many of whom
served for less than a year. Vicious factionalism accounts for some of the
shorter terms of office and the antipopes installed briefly in die Lateran, such
as Boniface VII or John XVI. A few popes removed from office, such as
Benedict VI, were murdered. Apart from the dominance in Rome itself of the
Theophylact and Crescentii families at various stages in the course of the tenth
centuries, other strands of political allegiance, to the houses of Wido of
Spoleto, Berengar of Friuli, die west Frankish rulers or the German kings,
influenced the choice of a pope. Yet diis did not always entail corruption and
disorder. John X, elected pope in 914, was a vigorous and experienced bishop
before his election and managed to form a political coalition of Italian rulers
against the Muslims 5 as well as fostering the chant school and the Lateran
administration during his fourteen-year reign. It was he who crowned
Berengar of Friuli, great grandson of Louis the Pious, as emperor in 915.
Under Count Alberic of Rome, moreover, there was a period of relative stabil-
ity from 932 to 954. Even diough the popes in this period, Leo VII, Stephen
VII (IX), Marinus II and Agapitus II, were Alberic’s nominees and well under
his thumb for the most part, much was achieved within the religious sphere,
notably in the reform of the monasteries of Rome and in ecclesiastical con-
tacts with Germany. From 1012 to 1044, moreover, there was effective leader-
ship from die papacy provided by die brothers Benedict VIII and John XIX
and their nephew Benedict IX. An indication of some popes’ aspirations to
emulate the early popes and the spiritual status of illustrious bishops of Rome
from the ‘pure’ days of its history lies in their decisions to adopt new names.
Silvester II, for example, chose his name in conjunction with the German
emperor Otto III in order to recall Constantine and his supposed relationship
with Pope Sylvester I. In the frequent choice of the names Leo, Gregory,
Damasus and Clement, in particular, earnest good intentions, if nothing more,
are mirrored.

The macabre affair of Formosus highlights die very specific objections that
could be mounted against a candidate for the papal throne, for it had repercus-
sions for nearly thirty years afterwards. Formosus, before his election to the
holy see in 891, had been an active missionary in Bulgaria, a papal legate and,
crucially, bishop of Porto, though he had fallen foul of John VIII and suffered
a period of excommunication. He proved to be an effective pope in his five
years’ incumbency, acting with astuteness in relation to England, Germany and
Constantinople and crowning Arnulf emperor in 896. But it was against canon

5 Zimmermann, Papsturkunden , no. 40, pp. 68—9.

Cambridge Histories Online © Cambridge University Press, 2008

140

ROSAMOND McKITTERICK

law for a bishop to be translated to another see. Never before had a bishop
become a pope and it was held against him, or used as the political excuse for
other actions, to an astonishing degree after his death. His successor Stephen
VI (VII) convened a synod and had the rotting corpse of Formosus exhumed,
dressed up in his pontifical robes, propped againt the papal throne and charged
with perjury, breaking canon law by being a bishop before he became pope and
coveting the papacy. Unsurprisingly, despite the oral defence offered by a luck-
less deacon assigned to the task of speaking on behalf of the corpse,
Formosus was found guilty and all his acts as pope declared null and void.
These of course included all his ordinations and consecrations, and, conven-
iently, Stephen Vi’s own disqualifying consecration as bishop of Anagni. The
body was then solemnly unfrocked and thrown into the Tiber. Stephen himself
did not last long after this gruesome outrage but was imprisoned and killed
soon afterwards. Partisans of Formosus subsequently elevated many of their
own candidates successively to the papal throne and the validity of Formosus’
acts was reconfirmed.

Too much emphasis on some lurid incidents, however, detracts from the
proper consideration of consistent elements of papal policy throughout the
tenth century and the reiteration of past agreements with the Carolingians
concerning their mutual obligations and the conduct of papal elections. At the
end of the seventh century it had still been necessary for the Byzantine
emperor to be sent a mandate sanctioning the consecration of die new pope
once the election had been held. The last pope to seek such sanction was
Gregory III (731-41), for the appointment of Zacharias (741-52) was made
without imperial ratification: all Zacharias did was to send an envoy to
announce his election and consecration. He in his turn was the last pope to
send formal notification of any kind to the eastern ruler, for Pope Paul I
(757-67) announced his election to Pippin III, king of the Franks and since 754
‘protector’ of the holy see. It is significant that Paul did so using the same
formula that had been used to notify the Byzantine emperors, except that he
did not ask for ratification, pledging instead undying loyalty. Similarly Stephen
III (IV) (768-77) dispatched an embasssy to the Frankish court to announce
his election, as did Leo III, who also sent Charlemagne the keys to the tomb of
St Peter and requested a Frankish envoy’s presence to receive the oaths of the
citizens of Rome. Thus the former ratificatory role of the Byzantine emperor
had been replaced by the pope feeling the need to announce his election to the
Frankish ruler, who was also the protector of Rome. Yet the balance shifted in
816 when Stephen IV (V) (816-17) anointed Louis the Pious. This was pre-
sumably an attempt to make the pope’s role in the creation of an emperor a
necessary one. In the reign of Paschal I (8 1 7-24), who was consecrated the day
after his election to pre-empt secular interference, the first major definition of

Cambridge Histories Online © Cambridge University Press, 2008

The Church

141

papal and imperial relations was set out in the document known as the Pactum
Ludovicianum (817). This confirmed the pope in the possession of his papal
states and patrimonies, and guaranteed the freedom of papal elections, though
it required the new pope to notify die emperor of his election and consecration
and to renew their treaty of friendship. Further, in 824, the Constitutio romana ,
ratified at a synod in Rome in 826, granted, among other tilings, immunity to all
under imperial or papal protection and restored the role of the people of
Rome in die election of the pope (suspended since 769). The pope was to
swear an oadi of loyalty to the emperor though papal independence in the spir-
itual sphere was maintained. It is from this date that we may observe the
resumption of local political interest in the papal elections and the re-
emergence of factions in Rome whose political ambitions were focused on the
papacy.

The Constitutio romana and the Tudovicianum were reconfirmed or adjusted
from time to time in the course of die tenth century. Thus the Constitutio romana
was revived in 898, to the extent that, while the papal election should be by the
bishops and clergy at the request of senate and people, papal consecration
should only take place in the presence of imperial emissaries. In 962, on the
occasion of Otto I’s coronation as emperor, die Ottonianum revived the
Tudovicianum. It confirmed the donations of Pippin and Charlemagne and
(possibly after December 963) restored the freedom of papal elections (subject
to imperial approval of the man elected and his obligation to swear an oath of
loyalty to the emperor).

Individuals also altered die relationship between die Frankish emperors and
the pope in practice, ei tlier making it even more extreme or asserting papal pre-
rogative. Thus, for example, Gregory IV (827-44) deferred his consecration
until the imperial legate had approved his election and he himself had sworn an
oath of loyalty to the emperor. Nicholas I was elected in the presence and with
the approval of die Emperor Louis II, but Sergius II’s consecration was rushed
through without waiting for imperial acknowledgement despite Lothar’s insis-
tence that a pope should not be consecrated except on his orders and in the
presence of his representative. Charles the Bald (emperor 875-7) refrained
from claiming a guiding hand in papal elections but Charles the Fat (emperor
881-5) insisted diat he ought to be consulted. From time to time in the tenth
century it was the pope who made the essential moves for die allocation of the
imperial tide to a potentially vigorous protector of Rome, but the emperor
himself could be called on to nominate a new pope. Thus when Otto III was
approached by the Roman nobility, he nominated Brun, son of the duke of
Carindiia, who took the name Gregory V and himself crowned Otto as
emperor and patrician. Similarly Henry Ill’s appointment as patrician enabled
him to take the lead in the appointment of the pope. Even at its lowest ebb the

Cambridge Histories Online © Cambridge University Press, 2008

142

ROSAMOND McKITTERICK

office of pope appears to have been more important than the person. This is
particularly the case as far as the coronation of Otto I as emperor in 962 was
concerned. The then pope was John XII, consecrated pope when possibly as
young as sixteen years old, and the son of die Count Alberic who had per-
suaded the Pope Agapitus II to accept John (then called Octavian) not only as
successor to Agapitus but also to Alberic, so that he would combine spiritual
and temporal rulership of Rome in one person. John however, if we are to
believe Liudprand of Cremona’s notoriously partial account, was renowned
for his dissolute life. Whatever the facts of the matter, John was arraigned at a
synod shordy after Otto’s coronation and actually deposed. Henry III, on the
other hand, appears to have wanted someone reputable to crown him emperor.
He deposed no less dian three popes at the synod of Sutri in 1046, before
installing Suiger, bishop of Bamberg as Pope Clement II. Clement crowned
Henry and Agnes his queen on 25 December 1046. It is significant, moreover,
that even the emperor’s men, such as Gregory V or Sylvester II, championed
papal prerogatives once safely consecrated. Their ability to do so was based on
that very stability within the papal administration mentioned earlier, as well as
the enduring attitude towards papal spiritual authority evident in die contacts
widi other countries. These are most obvious in die sphere of missionary
work, with the establishment of new bishoprics and the determination of dieir
ecclesiastical allegiances and liturgical observance in Denmark, Poland,
Hungary, Dalmatia and east of die Elbe. Sylvester II’s incumbency, for
example, saw the establishment of archbishoprics at Gniezno and Estergom
and the sending of a royal crown to King Stephen of Hungary. Occasionally,
theological issues as well as those of papal and episcopal jurisdiction in the
south of Italy were discussed and sides taken in disputes widi the eastern
emperor and the patriarch. Papal policy in its various contexts appears, ulti-
mately, to have been directed towards enhancing and confirming the authority
of the bishops of the Latin church.

Papal relations with the bishops of Latin Christendom are only a small part
of the concerns of the bishops themselves, however. Let us look at the careers
of some of these bishops and their clergy, therefore, and at the implications of
the written records of their activities. dEthelwold of Winchester, for example,
the celebrated reformer of the tenth-century English church, scholar, teacher
and ascetic, was elevated to the see of Winchester on 29 November 963.
According to his biographer, Wulfstan, he was an ‘intimate of the distin-
guished king Edgar’. He was able to temper the severity of his discipline with
coaxing gendeness, but was himself afflicted with frequent pain in his innards
and legs. 6 Wulfstan provides a lurid account of the scandalous and wicked

6 Wulfstan of Winchester, Vita sancti Aithelwoldi, cc. 25, 28, 30, pp. 42, 44, 46.

Cambridge Histories Online © Cambridge University Press, 2008

The Church

M 3

behaviour of the canons in possession at Winchester cathedral when
fEdielwold arrived. They were married, given to gourmandising and drunken-
ness, and some even did not celebrate Mass in due order. With permission
from King Edgar, dEthelwold expelled die canons, replaced them with
reformed monks from his own monastery at Abingdon, and thus became both
abbot and bishop. The canons’ livings presumably became the corporate pos-
session of the monks. The Life is loquacious on the reforming and monastic
practices of dEthelwold. He established Osgar as abbot at Abingdon in his
stead, installed nuns at Nunnaminster, and created a monastery at Ely under
Abbot Byrhtnoth and monasteries at Peterborough, Thorney and elsewhere.
All these were part of die monastic reorganisation throughout Europe
described by Wollasch, below, in which the Rule of Benedict was upheld as the
ideal. A remarkably successful attempt was made in England to impose
the Rule on ‘reformed’ monastic houses and on new foundations as well as the
English variation of imposing it on the cathedral clergy of the episcopal min-
sters.

The Vita sancti TEthelwoldi is conventional in its catalogue of dEthelwold’s
virtues. He is described as a consoler and helper of widows and orphans,
receiver of pilgrims and defender of the church. He refreshed the poor, and set
right those who had gone astray. As well as such obligatory pastoral concerns,
dEthelwold taught in the school at Winchester. He offered instruction in
grammar and translated Latin texts into English for the better understanding
of his pupils; many of his pupils became priests, abbots and notable bishops.
To complete the picture Wulfstan notes that a number of healing miracles are
associated with fEthelwold. Thus Wulfstan’s life of fEthelwold at a general level
provides a sympathetic account of late Anglo-Saxon religious devotion and
expression. Its focus on tire career of a saintly bishop, however, is typical of the
orientation of sympathy and perception of leadership in the church in Europe
in tire tenth and first half of the eleventh centuries expressed in the bishops’
vitae and collective histories of bishops in particular sees. If we compare
Wulfstan’s account of dEthelwold with those of other contemporary prelates,
there are both instructive contrasts and parallels, many of which raise issues of
concern to the history of the church as a whole.

Many of die German bishops, for example, accord well with the ideals set
out by dEthelwold, yet the career and Life of one in particular, exposes the
question of the criteria for saindiood and the recognition of holy status by the
ecclesiastical hierarchy. Ulrich of Augsburg was the first saint to be canonised
in what later became the normal fashion. It is a measure of his reputation for
piety that it was a petition from Augsburg itself in 993, exhorting Pope John
XV to acknowledge their former bishop as a saint, which achieved his
formal elevation to sainthood. Yet it may also be a measure of die increasing

Cambridge Histories Online © Cambridge University Press, 2008

144

ROSAMOND McKITTERICK

coherence of the ecclesiastical hierarchy and the role of the pope at its head (at
least in relation to the bishops of the Saxon empire) that such papal recogni-
tion was thought necessary to obtain. This is in striking contrast to die local
creation of saints and observance of their cults in earlier centuries. The I 'ita
sancti Oudalrici, written in about 982 by Gerhard, a priest ordained by Ulrich and
a member of the familia at Augsburg, played no small role in Ulrich’s recogni-
tion. In later years, dossiers on candidates for canonisation were presented to
the pope as a matter of course once Innocent III and his successors had
assumed control of the cult of saints by defining degrees of holiness and the
formal procedure for canonisation.

Gerhard related how the nobly born Ulrich conducted himself in his
diocese, with his rounds of episcopal visitation and exercise of pastoral care,
observance of die major liturgical feasts, the dedication of new churches, his
political service and loyalty to Otto I, especially during the quarrel widi Liudolf
and his uncle, Henry of Bavaria, his miracles, his journeys to Rome, his atten-
dance at the synod of Ingelheim in 972 and die high example, both in his own
conduct and in his attention to die discipline of odiers, 7 that he provided for
the conduct of the religious life. Some of diese features are clearly peculiar to
the circumstances of Ulrich alone, but odiers, such as his pastoral care, meticu-
lous ecclesiastical observance, personal piety and miracles, might be general-
ised as criteria for holiness.

It is die political dimension to Bishop Ulrich’s career above all, however, that
is a constantly recurring theme in the lives of the bishops of the tendi century.
Family politics, larger national tensions and the internal balance of authority
and power widiin the ecclesiastical hierarchy all play a role. Bernward, bishop
of Hildesheim (993-1022) is no exception. According to the I Ita sancti
Bernwardi (partly composed by Thangmar, who completed his portions
between 1022 and 1024, and partly much later towards the end of the twelfdi
century), Bernward served in his youdi as notary and was with the court in Italy
at the court of Otto II, and he and the see of Hildesheim itself were richly
rewarded by Otto. 8 Ecclesiastical prerogatives, such as Hildesheim’s jurisdic-
tion over the royal convent of Gandersheim, were jealously defended.
Bernward left other monuments to his incumbency, not least the magnificent
abbey church of St Michael, for the production of whose rich sculptural deco-
rations he was responsible, and the many liturgical manuscripts commissioned
by him or presented to him.

The German bishops, moreover, like their colleagues elsewhere in western
Europe, had much to do with the maintenance of educational provision in the
cathedral schools, actively promoting learning by their patronage of scholars

7 Gerhard, X^itasancti Oudalrici, c. 3, pp. 388— 90. 8 Thangmar, \ r ita sancti Bernwardi, c. 19, p. 767.

Cambridge Histories Online © Cambridge University Press, 2008

The Church

i45

and book production. Many boys, whether bound for ecclesiastical or for
secular careers, were sent for training to these schools - Trier, Augsburg,
Cologne, Eichstatt, Liege and Utrecht, under Bishops Ruotbert (931-56),
Ulrich, Brun, Ebrachar, Starchand and Balderic respectively, were particularly
celebrated. 9 Some visiting scholars, such as Stephen of Novara at Wurzburg,
provided an added attraction. In the Lotharingian sees of Metz and Toul, in die
Bavarian dioceses of Salzburg and Regensburg, and in the newly-founded
diocese of Magdeburg, schools flourished. Masters in these schools, such as
Ohtrich at Magdeburg, acquired great fame as scholars, and some bishops
clearly appointed notable scholars to add lustre to their schools. Under
Adaldag of Hamburg-Bremen, for example, the cathedral school was directed
by the learned master Thiadhelm. Most of the German bishops themselves,
moreover, either had been educated in one of these cathedral schools or came
from the schools of the reformed monasteries. Thus Hildiward of Halberstadt
and Balderic of Speyer had been educated at St Gallen. Otwin, educated at
Reichenau, promoted education as bishop at Hildesheim. They, and others like
them, came to their sees ready to apply in their new establishments what they
had learnt in their youth.

Many of these bishops were prominent not only in education but also in
their patronage of book and artefact production. Some bishops, such as
Theoderic of Metz, who in 984 donated a late eighth-century Homiliary
embellished with comments and corrections by Rather of Verona, 10 a histori-
cal miscellany compiled in the ninth century 1 1 and a ninth-century collection of
texts to do with computus and time 12 to the newly founded monastery of
Saint- Vincent at Metz, may have contented themselves with tire gifts of older
books. Other bishops commissioned new ones for their own or others’ use.
Archbishop Everger of Cologne (984-99), for example, gave an Epistle
Lectionary to die cathedral of Cologne, and had himself depicted therein in
abasement before the enthroned saints, Peter and Paul. 13 Gero, archbishop of
Cologne (969-76) had himself portrayed offering a Gospel Lectionary to St
Peter. 14 Perhaps the most famous German episcopal patron of book produc-
tion was Egbert, archbishop of Trier (977-93). The scriptorium of Trier
during his incumbency produced a number of books both for him and for
others (some of which were given to the cathedral at Trier), such as die famous

9 Anselm, Gesta episcoporum Leodiensium , c. 2 5 , p. 205.

10 Berlin, Stiftung Preussischer Kulturbesitz, MS Phillips 1676 (50).

11 Berlin, Stiftung Preussicher Kulturbesitz, MSS Phillips 1885 and 1896, and St Petersburg Saltykov-
Schedrin Library MSS Q.v.iv no. 5 and Q.v. class no. 9.

12 Berlin, Stiftung Preussischer Kulturbesitz, MS Phillipps 1831 (128).

13 Cologne, Erzbischofliche Diozesan- und Dombibliothek, MS Dom 143, fols. 3V— 4r.

14 Darmstadt, Hessische Landes- und Hochschulbibliothek, MS 1948.

Cambridge Histories Online © Cambridge University Press, 2008

146 ROSAMOND McKITTERICK

Registrum Gregorii with its verses lamenting die death of Otto II, 15 the Egbert
Psalter, 16 and the litde Greek and Latin Psalter designed to help the Empress
Theophanu learn Latin. 17 Yet die German bishops were not unique in their
patronage of scriptoria and the commissioning of books. Winchester also pro-
duced fine books for its bishops, notably the Benedictional of dEthelwold
(963-84) 18 and the Benedictional and Pontifical produced at Winchester c. 980
and subsequently associated with either Robert, archbishop of Rouen
(990-1037) or Robert of Jumieges, archbishop of Canterbury (1051— 5), of
which the former would appear to be the most likely first owner. Arnulf II,
archbishop of Milan (998-1018) had a handsome prayer book compiled and
decorated in gold, silver and rich colours at Milan for his personal use, of great
interest for the personal selection of prayers it contains. 19 The patronage of
the arts was not confined to books, however, as is clear from the reliquaries,
bronze sculptures, ivory-carvings and buildings also associated with Bishop
Bernward of Hildesheim (993-1022), another of the great German bishops
subsequently formally canonised.

It cannot be said that the tenth-century archbishops of Rlieims, on the other
hand, can be described in similar terms to those the German episcopal biogra-
phers and hagiographers invoke. Nor is it the case that the Rheims metropoli-
tans upheld the legacy of Hincmar in all respects. In their history there is a
tension between the ideal of Rlieims’ pre-eminence among the provinces of
the Frankish kingdoms and the reality of their unfitness for office, their politi-
cal intrigue with and subservience to members of the Carolingian family, mili-
tary prowess, immense wealth and the close, often oppressive association
between the archbishops and the principal monasteries of the diocese.
Heriveus (900-22), for example, led a force of 1,500 warriors to assist the
Frankish king against die Magyars; Seulf (922—5) strengthened die city’s fortifi-
cations. Not till 945 was the monastery of Saint-Remi finally able to obtain its
independence from archiepiscopal control and install a regular abbot. 20 Hugh
(925-31; 942-8), son of the Carolingian Count Heribert of Vermandois, was
first created archbishop at the age of five and only deposed, with royal inter-
vention, in favour of Artald (931-40), a monk of Saint-Remi. 21 It is small
wonder that Flodoard’s account stressed the physical actions and strength of
the archbishops, notably in building up the wealth of the archbishopric and
retrieving lost estates, at the expense of their spiritual leadership. The stress on
property should nevertheless be seen against such provisions as those of the
council of Trosly, summoned by Archbishop Heriveus in 909, which mounted
a vehement attack on laymen who had taken church land, and on the encroach-

15 Trier, Stadtbibliothek, MS 171a. 16 Cividale, Museo Archeologico, MSn.cxxxvi.

17 Trier, Stadtbibliothek, MS 7/9 8°. 18 BL, MS Additional 49 5 98.

19 BL, MS Egerton 3763. 20 Flodoard, HREiv, 32, pp. 583— 4. 21 Ibid., iv, 24, p.580.

Cambridge Histories Online © Cambridge University Press, 2008

The Church

*47

ment on ecclesiastical property, offices and privileges by the king and lay mag-
nates . 22 In the eyes of Flodoard, the retrieval of church land from such preda-
tors may indeed have appeared the highest episcopal virtue.

Flodoard had also given himself a much more difficult task than die simple
chronicling of one man’s life and career. In tackling the history of his see he
was attempting to provide not only a collective identity, but also a generalised
justification for the pattern of episcopal behaviour over many decades.
Ultimately, therefore, the identity and individual careers of the bishops are sub-
servient to the fortunes of the see of Rlieims itself, die real hero of his history.
He was able to outline a particular conception of the church in Frankish
society written in the form of a history and buttressed by copious citation of
legal documents which established Rlieims’ territorial rights. The concentra-
tion on property, however much it was at odds with fundamental other-worldly
Christian values, has the function of providing continuity and stability widiin a
changing world, beset by human intrigues and attack. Such collective identity
and emphasis on territorial possession, together with a firm conception of the
bishop’s role in relation to the lay world, are arguably the guiding motives in
many of the Carolingian, Saxon and Salian histories of bishoprics. The histo-
ries of Auxerre, Ravenna, Le Mans, Cambrai, Hamburg-Bremen, Naples,
Liege, Trier, Verdun, Metz and Toul were written in the wake of Paul the
Deacon’s Gesta episcoporum Mettensium of the late eighth century and inspired by
the example of the collective history of the popes, the Liber pontificalis. Their
distribution in itself may be significant as far as the existence of possibly differ-
ent perceptions of the bishops’ role in his diocese and in relation to his prede-
cessors and successors are concerned, for no histories in this genre appear to
have survived from Spain, England, northern Italy or southern France

Not until the incumbency of Adalbero could Rlieims again boast of an arch-
bishop who appears to conform more to the norms established in the episco-
pal vitae. Adalbero was from Lotharingia. His early education at Gorze bore
fruit in his own promotion of the school at Rlieims, for it was he who installed
Gerbert as master of the school at Rlieims, re-established a Benedictine com-
munity at Mouzon and made over many estates in the wine-producing areas of
the Meuse valley and Metz to the monks. A canon’s rule, moreover, was
reintroduced for the cathedral canons at Rlieims and regular monks were
restored to the monastery of Saint-Thierry. The scriptoria of Rlieims appear
to have been active at this stage as well. Adalbero at least, therefore, appears to
have resumed his predecessor Hincmar’s patronage of the cathedral library as
well as devoting himself to the embellishment of existing churches and
the erection of a new one in Rlieims itself . 23 Nevertheless, even taking the

22 Mansi, Concilia 1 8, cols. 263—308. 23 Richer, Historiae hi , 22—3, pp. 28/ 30.

Cambridge Histories Online © Cambridge University Press, 2008

148

ROSAMOND McKITTERICK

presentation of Adalbero in Richer’s Historiae into account, the political
involvement of the archbishops of Rheims remained secular in character.
Artald’s advocacy in the takeover of the west Frankish kingdom by Hugh
Capet, for example, is well attested. 24 But then, it was in Rheims’ interests to
assert to posterity its enabling and crucially supportive role in relation to the
crown. Unlike tire bishops in England, however, the archbishops of Rheims
made little attempt to exert moral authority over their rulers, and the
Carolingian tradition of self-seeking ambition and political interference was
staunchly maintained by Adalbero ’s successors.

A similar emphasis on the physical strength of the bishops can be discerned
in tire meagre account from Sens. Sens under Archbishop Archembald
(958-67) had been a disgrace, even if measured against the criteria implied by
Flodoard of Rheims. According to the eleventh-century chronicler of Sens,
Odorannus (and Clarius, who echoes him), Archembald had actually sold not
only the lands and churches of Sens but the actual church buildings them-
selves. He spent the proceeds on self-indulgence, turning the refectory of St
Peter’s monastery into a brothel and keeping his hunting dogs in the monastic
precincts. Under Anastasius (968-76) and particularly under Seguin (977-99),
however, the church of Sens was restored to the position of respect it had
attained in the ninth century. Such rapid recovery endorses Fichtenau’s less
censorious view of the effect of temporary alienations. The chroniclers are
most interested in Seguin’s devotion to the monastery of St Peter, and give full
details of his restoration of die discipline, the fabric and the estates of the
monastery and the installation of a new abbot. Seguin also had the cathedral of
St Stephen, which had been destroyed by fire in July 967, rebuilt, and he and his
fellow bishops of Troyes, Nevers and Auxerre presided over its reconsecra-
tion. The needs, of the laity received some attention in the establishment of
parish churches in the city and the acquisition of important new relics, includ-
ing die arm of Pope Leo the Great, to act as a focus of devotion. Both Seguin
and Leothericus his successor are reported to have received the pallium from
the pope. Although this was the normal recognition received by a new arch-
bishop since at least tire eighth century, Odorannus insists that this pallium also
conferred the primacy of Gaul upon the archbishop of Sens. Whether this was
really the case is doubtful. Rheims too claimed the primacy, yet the precise
function of this role, as distinct from its honorific nature, even in die days of
Boniface of Mainz and Chrodegang of Metz, is difficult to determine.

Whereas the bishops of Rheims and Sens worked within an ancient eccle-
siastical framework in their efforts to cling to their pre-eminence, the archbish-
ops of Hamburg-Bremen, associated as one bishopric from 864, attempted to

24 Ibid., iv, 3— 5, pp. 128/32.

Cambridge Histories Online © Cambridge University Press, 2008

The Church

149

exert their ecclesiastical authority in Denmark, Norway and Sweden as well as
among the Slavs east of the Elbe. Ever since their first bishop, Anskar, had
preached the Gospel in Denmark and Sweden, die bishops, in the way Adam of
Bremen tells their story, directed their energies towards establishing the church
in Scandinavia. In particular, Unni (bishop 919-36) and a ‘very holy man’,
preached and died in Sweden; Adaldag established the dioceses of Ribe,
Schleswig and Aarhus under die metropolitan authority of Hamburg-Bremen
and also preached among the Slavs. The bishops worked in such a way as to
consolidate die archbishops’ own ecclesiastical authority widiin the northern
part of the east Frankish kingdom. Their missionary work in the nordi, more-
over, cannot be divorced from political considerations. A constant, but under-
played, presence in Adam’s story of the northern mission is die clashes of
interest with England, particularly after the Danish conquest of England and
during the rule of Cnut. 25 Similarly, in die overtures to the Slavs, political and
ecclesiastical expansion are closely associated, and the lack of Frankish and
Saxon success among the northern Slavs is due in part to the peoples of that
region’s staunch efforts to retain dieir political autonomy.

Despite the wider horizons of diese northern bishops, diey exemplify many
of die customary episcopal virtues. Thus Adaldag distributed among the par-
ishes of his diocese the relics of the holy martyrs (Quiriacus, and Cesarius,
Victor and Corona, Felix, Felician, Cosmas and Damian) and saw to the main-
tenance of the xenodochium of Bremen. Adaldag himself, moreover, was from
Flildesheim, ‘noble in appearance and behaviour, illustrious in family’ and
related to Adalward of Verden. He had served Otto I in his chancery. He
founded a convent at Hieslingen and a monastery at Reepsholt and maintained
a school at Hamburg. Unwan, bishop 1013-29, was similarly of good family,
selected from Paderborn, rich and generous. He was the first to impose die rule
of canons on the cathedral clergy and did all he could, by loading diem widi
gifts, paid for out of the diocesan treasury of Hamburg, to soften up the kings
of Norway towards die ecclesiastical overtures from the German church rep-
resented by Hamburg-Bremen.

Adam is more dian a biographer of the separate bishops who are his sub-
jects. He provides nothing less than a history of his region in which Hamburg-
Bremen is the focus; it is a world where the Saxon kings and the rulers of
the Slavs and of the Scandinavian peoples all interact in the political and
secular concerns of the archbishops. Aldiough the mission to the headien is
the mainspring of the account, Adam contrives to convey the principles on
which political and ecclesiastical expansion can be based. He sets the agenda
at the beginning by his unequivocal account of the career of Anskar and its

25 Adam of Bremen, Gesta u, 37 , p. 98 .

Cambridge Histories Online © Cambridge University Press, 2008

150 ROSAMOND McKITTERICK

significance. He then cuts the careers of the later bishops to the cloth of
Anskar. Devotion to Hamburg-Bremen is the highest virtue, and the arch*
bishop who earns the most censure is Hermann (1032—5). Hermann was
elected from the chapter of Halberstadt. He rarely visited Hamburg except to
come with an army and lay it waste. His one virtue was that he installed Guido
as music master in the cathedral at Bremen, who reformed chant and liturgical
discipline.

The archbishops of Hamburg-Bremen were not alone in die tenth century
as far as expansion of the frontiers of Christendom was concerned. Otto I had
extended the frontiers of his kingdom towards the east, and establishment of
the church with die assistance of his bishops within diat region was a funda-
mental component of political consolidation. Cnut and the English bishops
had, as already noted, mounted an independent missionary enterprise in
Scandinavia that clashed with die interests of die archbishops of Hamburg-
Bremen. Missionary endeavour elsewhere is signalled on the one hand by the
foundation of new bishoprics, such as Bamberg (1007), Magdeburg (968),
Gnesen ( 999), Posen ( 968) or Prague( 973), which were intended to act as mis-
sionary outposts and new ecclesiastical centres, and on the other by the cele-
bratory lives of the saints who brought Christianity to the Danes, Slavs,
Obodrites, Rus', Poles or Magyars. In their careers, and in the first introduction
of Christianity in however superficial a manner, the princes who first accepted
Christianity often played a key role, as did Wenceslas, Boleslav I and Boleslav II
in Bohemia, Miesco I in Poland, Harald Bluetooth (950-86) in Denmark and
Olaf Tryggvason (995-1000) and Olaf Haraldson (101 5-1030) in Norway.
Recent research, indeed, has tended to emphasise die role of the prince and the
leaders in any region rather than that of the foreign missionaries in die decision
to adopt Christianity. That the new churches established were essentially state
churches certainly supports this view. Conversion to Christianity, therefore,
was a decision not merely about religion but also about political association and
cultural alignment. Such religious commitment could also lend coherence to
different groups of poeple, or enable an individual to consolidate political
control, as in die case of Vladimir of Kiev, or Miesco of Poland. In odier
words, the stabilisation of royal power and Christianisation often went hand in
hand. Missions sent out from Salzburg and Aquileia in the eighth and ninth
centuries were built upon, and in a number of regions either Italian churches
(or even die pope himself) or the Byzantine church played a prominent role ini-
tially. Moravia was Christianised, for example, by the Byzantine missionary
Methodius (commemorated now in die series of wall-paintings in the lower
basilica of S. Clemente in Rome). Rorivoj and his wife Ludmila were baptised
together with their son Spytihnev by archbishop Methodius at the court of
Sviatopolk. In 895, however, Spytihnev renewed Moravia’s allegiance with the

Cambridge Histories Online © Cambridge University Press, 2008

The Church

1 5 1

Franks and it has been conjectured that the jurisdiction of the bishop of
Regensburg was extended into Bohemia. In the late tenth century a bishopric
of Prague was created whose second bishop was Adalbert Vojtech. He had
been educated at Magdeburg. He was canonised in 999, and thus gave the
Bohemians a national saint, soon to be joined by King Wenceslas. Croatia was
reunified under Tomislav (910-29), who was recognised by die pope, for what
that was worth, as king of die Croats. For strategic reasons, Venice, Byzantium
and Hungary were all interested in the kingdom of the Croats. Thus in this
region, political and ecclesiastical interests clashed, for the archbishop of Split
aspired to ecclesiastical dominance of the region.

It is difficult to determine the nature and strength of the paganism of these
converted peoples; we only hear about them from the Christians who con-
verted them, and who resort to facile judgements and contemptuous refer-
ences about the religious beliefs and practices of the pagans. Some notion of
the strength of the paganism of the eastern Slavs in particular, however, might
be surmised from the long resistance of the Elbe Slavs to conversion. Further
east, the Lidiuanians were not converted to Christianity until the mid-four-
teenth century, and then, apparendy, more as a political move dian from relig-
ious conviction. In some cases, moreover, the paganism of some of these
eastern regions may have been watered down by long-standing contacts with
Christian regions. Resistance to Christianity, moreover, was part and parcel of
political resistance. It is significant, for example, that die conversion of
Norway was as slow as it was. Although missionaries had been present in the
country since the end of the ninth century, ecclesiastical structures did not
stabilise for another 200 years. A change in religion transformed die life of the
community, though in Iceland, on the odier hand, striking adaptations of
Christianity to Icelandic society are apparent. The Christian church had to be
accommodated physically on the land and socially within the local commu-
nities. In the Slav and Scandinavian regions the accommodation to and accep-
tance of Christianity required die decision not only of the ruler but also of his
people. The communal nature of the decision-making is expressed symboli-
cally by die famous Althing of Iceland in 999—1000 when Christianity was
declared to be the official religion of Iceland.

It is indicative of die power of the bishops that it was largely through osten-
sibly Christian means that political ends were achieved. Rather than the careers
of founders of monasteries and ascetic monastic saints predominating in die
hagiography, as is the case before die tenth century, it is the bishops, sometimes
ascetic, sometimes also founders of monasteries and promoters of monasti-
cism and displaying many of the attributes of the saints of the earlier middle
ages, as a result in part of earnest efforts on the part of their hagiographers
to make diem so. In addition, however, they are always practical, involved in

Cambridge Histories Online © Cambridge University Press, 2008

I 5 2

ROSAMOND McKITTERICK

politics, energetic, learned, and interested in die material and physical welfare
of their flocks. The focus of these bishops’ lives, according to the accounts
given of them, is, as we have seen from the examples above, essentially local.
The fact that we see these bishops primarily through the eyes of their hagiogra-
phers might be thought to diminish dieir credibility as historical evidence, were
it not for dieir indirect display of contemporary expectations of a bishop.
Bishop Ulrich, or Bishop fEthelwold, and the others at die hands of their biog-
raphers, dierefore, each provide a speculum episcopi, a model for other bishops to
follow and against which laity and lower clergy alike could measure their own
bishops. Nevertheless, like much verbal, written image-making, the portraits
we have of these bishops are indeed determined in part by their own activities
and personalities in reality and in part by the activities of dieir predecessors, as
we saw in die case of the bishops of Hamburg.

More can be learnt of many of these bishops in their own words, in the form
of letters exchanged by many of them, sometimes, as in the case of Gerbert of
Rheims and Rather of Verona, in copious quantities. From these can be
observed the entire gamut of activity on the part of bishops, from political
intrigue to pastoral advice. Further, we see these bishops all over Europe par-
ticipating in die synods of their day, and it is these which establish the degree to
which they felt they had a collective identity and acted corporately. A striking
feature of much of the ecclesiastical legislation of die tenth and early eleventh
centuries, however, is its local and diocesan nature. Very few general or even
regional synods appear to have been convened, in marked contrast to the
Carolingian period. Rather, the decrees preserved are more often from dioce-
san, or at best provincial meetings. Nevertheless, they reflect a unity of
purpose and similarity of preoccupation among church leaders across the
whole of Europe, given local differences. Despite its apparently normative
character, moreover, the synodal legislation of die church was usually a
response to particular problems and setded local disputes as well as occasion-
ally issuing directives on general religious observance and organisation. A
remarkably large quantity of such material survives, albeit it is often difficult to
categorise. Some documents, for example, are classified as conciliar proceed-
ings, when they read more like the records of an ecclesiastical court. Others are
admonitory or exhortatory in character, or set out to prescribe remedies for a
particular set of abuses. Rather of Verona, for example, apparently chose the
forum of a synod to address his clergy on their duties, especially celebrating the
Mass, the need to teach all parishioners the Lord’s Prayer and Creed, the proper
days for baptism (Easter Eve, die eve of Pentecost) and what the priest himself
should know: the Mass, the Gospels and Episdes, and the rites for die sick, the
dying and the dead . 26

26 Rather of Verona, Epistolae, ed. Weigle, no. 25 , pp. 1 24—37.

Cambridge Histories Online © Cambridge University Press, 2008

The Church

1 5 3

Sometimes a synod would be convened as a consequence of a visit from an
external ecclesiastical dignitary, such as the synod of Hohenaltheim in 916,
prompted by the visit of the papal legate Peter of Orte. Others settled disputes
over the appointment of particular bishops, such as the clutch of meetings
held at Verdun, Mouzon, Ingelheim, Trier and Rome precipitated by the
scandal of die archbishopric of Rheims between 947 and 949, the Cologne
provincial synod at Cologne in 920 concerning the quarrel over the bishopric
of Liege, or the synods of Duisburg (929), Ravenna (9 5 5) and Ingelheim (958).
Provincial and diocesan synods in particular discussed general internal eccle-
siastical provisions relating to die behaviour of the bishops and clergy, and
matters of church discipline or of the religious observance of the laity. The
presence of Italian bishops at meetings in Germany from the 950s indicates
the extra dimension given ecclesiastical politics by Ottoman involvement in
Italy. Many of these synods, particularly those concerning disputes over sees,
had an obvious political dimension, though only rarely were synods actually
summoned by a king. Those drat were, were called on die initiative of such
rulers as Henry I or Otto I in such a way as to retiect the king’s close interest in
the day-to-day running of ecclesiastical affairs and die degree to which this was
associated by die king with general good government. One of these, the synod
of Erfurt (932), may shed light on the ecclesiastical policy of Henry I, for it is
possible that the king had concerned himself intimately with both its agenda
and its organisation. Thus, like his Carolingian predecessors, Henry would be,
on this reading, implicitly identifying his expectations as ruler with those of his
bishops as pastoral and spiritual leaders of their dioceses. His attendance in
person endorsed his royal position in relation to the church. The synod of
Erfurt in its turn was enhanced by the king’s presence and was attended by the
archbishops of Mainz and Hamburg-Bremen and the bishops of Verden,
Strasbourg, Constance, Paderborn, Augsburg, Halberstadt, Wurzburg,
Osnabriick, Munster and Minden, as well as a host of abbots and other clergy.
In their deliberations, reference was made to the earlier councils of Mainz (852)
and Tribur (895), which in their turn reflected earlier Carolingian provisions.
The conciliar decisions of the tenth century overall have a clear and acknowl-
edged debt to Carolingian church councils.

This raises the issues of the context in which the records of the proceedings
have survived, the development of canon law, the definition and understand-
ing of the authority of the church, and the role of historical precedent. The
current understanding of the earlier tenth century, rightly or wrongly, is as one
of transition between the concerted activity of the Carolingian synods and the
certainties of eleventh-century Gregorian ecclesiology. Although there are not
so many supra-regional councils as in the ninth century, there is certainly a
respectable number of local and provincial assemblies in the tenth century,

Cambridge Histories Online © Cambridge University Press, 2008

154

ROSAMOND McKITTERICK

especially in die east Frankish kingdom, with records surviving, and many
more, of which we have at least notices announcing that diey had been con-
vened. As a form of church government and decision-making, the synod was
undoubtedly regarded as appropriate. Not only were many Carolingian concil-
iar decrees incorporated into collections of canon law in the late ninth and the
tenth centuries, Carolingian decisions, as has been noted, are also specifically
recalled in the deliberations of tenth-century and later ecclesiastical assem-
blies. If continuities can be so clearly observed, however, it throws doubt on
the usefulness of the concept of transition. It suggests, moreover, diat die late
Carolingian bishops and early Ottoman bishops were, practically speaking,
indistinguishable from one anodier in terms of the traditions and practice they
acknowledged.

Certainly from a west Frankish perspective die preoccupations of
Carolingian ecclesiastical legislation remained current issues in the context of
tenth-century ecclesiastical deliberations. The general concerns of the few sur-
viving decrees of synods between 888 and 987 are, for die most part, indistin-
guishable from those of their ninth-century precursors, even if particular local
and provincial synods can be seen addressing immediate problems. As men-
tioned above, the synod of Trosly in 909, for example, forsook generalities in
its specific attack on laymen who had taken church land and the encroachment
of ecclesiastical property, offices and privileges by the king and lay magnates
which the tenth-century bishops of Rlieims did their utmost to counteract.

Yet the emphasis on discipline and the moral underpinning of the church,
on episcopal and priestly accountability, and on the maintenance of the organ-
isation of the church as part of the Christian realm remain the same as in the
heyday of die Carolingian councils under Charlemagne and Louis the Pious. In
the manuscript traditions of the great councils of the Carolingians in west
Frankish sources, moreover, we get exactly die same picture as from the east. It
is clear from the context in which many Carolingian conciliar decrees have sur-
vived, that individual tenth-century bishops made a direct connection between
the preoccupations and concerns of die Carolingian synods and their own. A
tenth-century manuscript from Freising, for example, contains the decrees of
the Carolingian reform councils in company with the decrees of
Hohenaldieim (916), Koblenz (922), Duisburg (929) and Erfurt (9 3 2). 27 The
compiler, possibly under the auspices of Bishop Abraham of Freising, pre-
sumably wished to make a point about the relationship between die two sets of
decrees, a century apart in date but not in aspiration. Further, many of these
synodal decrees, together widi the major conciliar canons of the ninth century,
were, in dieir turn, incorporated into such major elevendi-century collections

27 Clm 27246.

Cambridge Histories Online © Cambridge University Press, 2008

The Church

of canon law as that of Burchard of Worms. Older Frankish compilations of
canon law such as the Vetus gallicavrete also widely spread throughout western
Europe from the early eighth century onwards and were still current in collec-
tions of the tenth and eleventh centuries. The Collectio canonum hibernensium
continued to be popular, as did the Concordia Cresconii (in Italy) and the so-called
Roman collection, the Dionysio-Hadriana, and Collectio Hispana. Further compi-
lations drawing on these were consistently compiled in all kinds of contexts.
Added to these, of course, was die collection of notorious forgeries known by
us as the Pseudo-Isidorean Decretals, though many compilers in the tenth and
early eleventh centuries appear to have drawn on both genuine and forged
canon law. A notable example is the collection attributed to Remedius of Chur,
probably produced in southern Germany c. 900 (or 870 according to Reynolds)
and found in many German dioceses, and the Collectio Anselmo dedicata, dedi-
cated to Bishop Anselm II of Milan (882- 96) which enjoyed fairly wide circu-
lation in Italy in the tenth century.

Obviously, each of the books containing a kaleidoscopic variety of different
canons from Carolingian and earlier church councils and canon law collections
was compiled for a particular purpose and at a particular time. Some may well
have been in relation to the agenda for particular synods or in relation to major
disputes. One problem, however, is how the records of a synod were actually
made. A few extant tenth-century charters from the diocese of Sens, for
instance, indicate that the initial record of a synod was in a form resembling
notitia from a court case, with decisions made and those present recorded. All
such documents imply an agreement by the participants to uphold what had
been decided, but there is no intrinsic reason why such decisions should have
continued to carry authority once they were translated out of their formal dip-
lomatic context and inserted into a collection of legal material compiled for
other purposes. Yet a further consideration is that every synod that met to
discuss matters to do with the organisation of the church, belief, discipline or
specific and immediate disputes placed itself, whether implicitly or explicitly, in
a historical succession. The compiler of a collection of ecclesiastical legislation
c. 1000 within the province of Mainz, for example, made an explicit link
between the early Christian church, Visigothic, Anglo-Saxon and Carolingian
decisions and tire church of his own day by intermingling synodal decrees from
all these regions. 28 The clauses relate to the practical and common concerns of
the church, such as the role of the priests, the jurisdiction of bishops, the inci-
dence of superstition, problems of valid marriages and what priests should
preach. A west Frankish collection compiled for the use of a bishop includes
among other texts the capitula of Riculf of Soissons, ordines for the convening

28 Wolfenbiittel, Herzog- August Bibliothek, MS Helmst 454 .

Cambridge Histories Online © Cambridge University Press, 2008

1 56

ROSAMOND McKITTERICK

of synods and much to do with the proper conduct and duties of priests and
the relations between a bishop and an abbot . 29 Some collections, such as that of
Burchard of Worms or Regino of Prum, acquired greater authority for reasons
that can no longer be established, though their comprehensiveness may have
made them particularly attractive to later compilers. One function of these col-
lections, therefore, became that of affirming ecclesiastical and episcopal
authority. The individual selections made may also indicate what one person
had decided was important from die past legislation at his disposal. A collec-
tion compiled for the archbishop of Mainz in the second half of the tenth
century, for example, takes as its core Regino of Priim’s De synodalibus causis et
disciplinis ecclesiasticis and organises round it various decisions concerning the
administration of die church, canon law and ecclesiastical organisation, all of
which is direcdy related to the work of the bishop in his own diocese . 30
Similarly, many bishops, such as Ruotger of Trier, continued the ninth-century
episcopal practice of issuing capitula or sets of directives to the clergy of dieir
dioceses. They mirror diocesan concerns and are especially geared to the day-
to-day running of parish affairs, the conduct of the clergy and the faith of the
laity.

Faith, coupled with the total reorientation of the life of the laity diroughout
the Christian year, was expressed in the liturgical rituals of die Christian
church, and some of the most creative episcopal activity of the tenth century
was in the sphere of liturgy and liturgical chant. The main liturgical books of
the early middle ages comprised die sacramentary, containing the texts for the
celebrant of the eucharistic prayers and prayers for other rites of the church
throughout the liturgical year, and the ordines , containing descriptions of and
instructions for the rituals of various kinds at which the prayers in the sacra-
mentary were to be recited. Thus we can reconstruct how the liturgy was actu-
ally celebrated as well as what was said. Individual bishops (and abbots) were
responsible for an intensely local adaptation of the liturgical books and chant
repertoire. Thus every surviving liturgical codex provides some slight variation
and was designed for use in a particular church or group of churches. In addi-
tion, the plethora of commentaries on the Mass produced in the ninth century
continued to circulate. Despite efforts to attain liturgical uniformity, indeed,
die variety of rite throughout Europe by the end of the ninth century was, if
anything, even greater than it had been two centuries earlier, with indigenous
traditions heartily preserved. The liturgical evidence from the tenth and early
eleventh centuries, moreover, illustrates how this great local diversity was
maintained and augmented in an extraordinarily creative period. First, there is
the great range of manuscripts extant, from centres as diverse as Regensburg,

29 BNlat. 4280A. 30 Wolfenbiittel, Herzog August Bibliothek, MS 8321 Aug.2 0 .

Cambridge Histories Online © Cambridge University Press, 2008

The Church

1 57

Mainz, Basel, Milan, York, Winchester, St German’s in Cornwall, Cracow and
Sens. Secondly there was the creation of a new type of composite book, a
Missal, which contained mass texts, Episdes, Gospels and Antiphons, though
other separate books were created such as Benedictionals, that is, those con-
taining episcopal prayers. Pope Gregory V (996-9) himself requested the mon-
astery of Reichenau to send a Missal for papal use in 998. Further, a new book
for bishops, a Pontifical, was devised. It combined prayer texts, rites and ordines
for a bishop in one volume. The most famous version is the so-called Romano-
Germanic Pontifical, put together at Mainz under the archbishops (Frederick,
William and Willigis), in the mid-tenth century, incorporating much older
Frankish non-eucharistic material, notably from the Hadrianum Sacramentary
as supplemented by Benedict of Aniane for Charlemagne, with which the
ordines were combined. 31 Of particular importance for this was a ninth-century
compilation from St Gallen. 32 The Romano-Germanic Pontifical rapidly
superseded all other collections of ordines throughout the huge dioceses of
Mainz and Salzburg, though its influence elsewhere in Europe was more
limited. It is in the ordines, even more than in the texts of the prayers, that die
innovations introduced by the Franks may be fully appreciated. They clearly
reflect local episcopal initiatives as well as die bishops’ ability to have recourse
to scriptoria equipped to supply liturgical codices as needed.

Thus in the east Frankish kingdom between 880 and 900 the ritual descrip-
tion of an episcopal Mass ( Ordo ix) was added to the ordines and incorporated
into the Mainz Pontifical. Burchard of Worms devised a composite list of Old
and New Testament readings for the night office ( Ordo xmc) and a ceremonial
description was provided for liturgical functions throughout die Christian year
( Ordo l). Other ordines reflect west Frankish innovations, such as the ceremonial
description of the last three days of Holy Week from Saint-Martial of Limoges
of the late tenth century ( Ordo xxxm). 33 In Rome c. 925 was added an ordo for
the ordination of lectors, acolytes, subdeacons, deacons, priests and bishops
{Ordo xxxv) and other ordines for episcopal consecration were added in about
970 {Ordo xxva and xxxvb). A rite for the coronation of an emperor was used
between the death of Charles the Fat and the coronation of Berengar in 91 5
{Ordoxui).

Liturgical uniformity is thus not in evidence. Even when new collections
were formed, as in the tenth century, these were from the older familiar materi-
als with particular liturgical allegiances displayed. An obvious example is the
rites of the different regions brought within Christendom in the course of die
tenth century - Croatia, Hungary, Bohemia and Poland. The claims of

31 See, for example, BN lat. 1 3 3 1 3 and Vienna, Osterreichische Nationalbibliothek, MS lat. 701.

32 St Gallen, Stiftsbibliothek, MSS 614 and 140. 33 BN lat. 1248.

Cambridge Histories Online © Cambridge University Press, 2008

158 ROSAMOND McKITTERICK

Byzantium and Rome appear to have been equally enticing, but the liturgical
orientation, Latin in Bohemia, Slavic in Moravia, Glagolitic in Croatia, Latin in
Hungary, reflected a cultural, linguistic, intellectual and political orientation of
great significance for the future development of these regions. Choice of
which book to adopt in any church clearly rested with the individual bishop.
For use in the Aachen palace chapel in the time of the Emperor Lothar, for
example, it appears that an earlier stage of Gregorian Sacramentary which had
been a Roman priest’s Mass Book (as distinct from one for use by a pope like
the Hadrian urn) was used by the scribes of the ‘court’ scriptorium at Aachen. 34
It was subsequently used at Liege, and migrated to Verona between 900 and
950 where further additions were made to it. Tenth-century examples of the
eighth-century Gelasian demonstrate how older Frankish liturgical texts were
still in use. 35 In regions but recently brought under Frankish rule there is a pre-
dominance of ‘non-Roman’ rites such as the ‘Celtic rite’ in use in Ireland, the
Ambrosian or Milanese rite which remained current in parts of northern Italy
even into the eleventh century, and the ‘Mozarabic’ and ‘Old Spanish’ liturgies.
The latter is represented, for example, by the reorganisation of die Mozarabic
Antiphonary associated with Bishop Akilia of Leon (917— 70). 36 All, neverthe-
less, played some role in the compilations of the Frankish kingdoms as well,
and cannot themselves be said to be completely separate from the Roman-
Frankish texts.

Other books were those connected with liturgical chant, such as the
Responsaries, Tropars, Sequentiaries and Hymnaries or Hymnals, and those
containing texts used during the Mass or in other ceremonies, such as the epis-
copal benedictionals and martyrologies. It is the lectionaries which demon-
strate the greatest local diversity in the arrangements and selection of texts in
the early middle ages (apart from die Spanish ones which were relatively
stable), though when the Roman Catholic church fixed the readings in 1 570, it
was an essentially Romano-Frankish system that was chosen.

The most dramatic additions and changes wrought to the chant books, on
the other hand, were in relation to their music. In the late Merovingian and
Carolingian periods, a hybrid ‘Roman’ chant repertoire had been created by
mixing older material of Roman origin with earlier indigenous Frankish
material. This in its turn was combined with what was understood or claimed
by the Carolingians to be contemporary Roman music, in order to create a dis-
tinctive liturgical chant commonly known as ‘Gregorian’. The tenth century
continued the remarkable proliferation of musical notations and the expan-

34 Padua, Biblioteca capitolare, MS D47.

35 Zurich, Zentralbibliothek, MS C43, made 1020—30 and the so-called ‘Missal of Monza’, Monza,

Biblioteca capitolare, Codex fi— i— i from Bergamo, of the late ninth or early tenth century.

36 Leon, Biblioteca capitolare, MS i .

Cambridge Histories Online © Cambridge University Press, 2008

The Church 1 5 9

sion of the chant repertory begun in the ninth century, especially of the new
syntheses of melody and prose known as sequences and tropes particularly
associated with such centres as Liege, Reichenau, St Gallen, Winchester, and
many cathedrals and monasteries in the west Frankish kingdom and in Italy
after about 1 000. There were also further developments (notably on the part of
Hermann of the Reichenau and Guido of Arezzo) in chant theory, the wider
use of such musical instruments as die organ, and die emergence of liturgical
plays in the late tendi century.

The liturgical year, celebrated in bodi words and music, included not only
the great Christian festivals but also die feasts of local saints. Recent discus-
sions of popular religion and culture in the middle ages have accorded promi-
nence to the evidence of die saints’ Lives, and other information concerning
the observance of dieir cults. In the light of such Lives of saints as the I Ita
sancti Oudalrici, however, and the indications of the importance of historical
context, it is clear that saints’ Lives cannot be regarded as a static genre provid-
ing a stable category of historical evidence throughout the early middle ages.
The old distinctions between official liturgical cults with written texts and
popular oral veneration of saints (thought to be proof of quasi-pagan super-
stition) are no longer valid. Recent work has stressed how both hagiography
and the observance of a cult are far from indifferent to chronology and histori-
cal context and that the ‘popular’ veneration of a saint need not involve non-
Christian, magical or pagan practices. In die diocese or Orleans, for instance,
the cult of long-dead local fadiers of the Orleannais formed the focal point of
the cult of saints in die region. Local models of sanctity were provided. Thus,
the saints’ role in local communities is to be understood within die context of
both spiritual and social beliefs; die patronage exerted by a saint was under-
stood in terms of the existing socal system. The ‘logic of saindy patronage’ has
been invoked by Head to explain the way in which laity and clergy interacted.
The duties of patronage required the saint to act on behalf of his or her ser-
vants; the saint, with his or her miraculous powers, would protect those who
chose to become his or her servants, and this relationship in many ways mir-
rored those of lords and fidelesi n secular society. The submission of a person to
the power of a saint and the request that the power be used to protect and
intercede for die petitioner governed the full array of relationships between
the local ‘father’ and his servants. God acted through the saint. Saints were live
presences who owned property, appeared in visions, cured die sick, meted out
punishment and dispensed justice. Miraculous powers were exhibited in their
relics. The accounts of their lives provided a means of observing and inter-
preting die relics. The observance of the Orleannais saints was first recom-
mended by Bishop Walther of Orleans in his statute of 871. In the tenth and
eleventh centuries, accounts of these lives were written in order to strengthen

Cambridge Histories Online © Cambridge University Press, 2008

160 ROSAMOND McKITTERICK

and particularise the saints’ associations with the Orleannais region. Evutius,
Anianus, Benedict, Maximinus, Lifardus and the relic of tire Holy Cross treas-
ured at the cathedral of Orleans were firmly associated with a particular relig-
ious community and the locality within which it was situated. The relics of
other saints - Paul Aurelian and Maurus the martyr - were translated into the
diocese in the course of the tenth century. The Orleannais vitae, like other early
medieval saints’ Lives, had both a private and a public function. That is, they
were intended for both tire legentes and tire audientia populi, and thus were of
crucial importance in nourishing the popular devotion of the laity. The cult of
saints functioned in many communities all over Europe just as it did in the
Orleannais. Despite the fact that hagiography was a well-established literary
genre with clear formal conventions and a standard repertoire of stylistic
models, the author of each vita managed to tailor his material to his own spe-
cific requirements, in which immediate and perennial concerns were balanced
and images of sanctity modifed to suit the preoccupations of a particular audi-
ence. Lay patrons commissioned vitae and many laymen and laywomen were
celebrated as saints.

The cult of tire Virgin Mary in later Anglo-Saxon England, on the other
hand, reveals the extraordinarily developed cult of a universal saint, especially
in association with the reform movement of the tenth century. Mary is com-
memorated in a large number of dedications of churches and monasteries, in
the composition of private and monastic prayers, tire celebration of Marian
feasts throughout the liturgical year in ecclesiastical communities, the acquisi-
tion, from the tenth and eleventh centuries, of such relics of Mary as frag-
ments of her clothing, hair and sepulchre and some of her milk, her portrayal
in art and the writing of texts, especially homilies in Old English describing her
life and death.

Relics of the Virgin Mary abound on the continent as well and statues of her
as well as of other saints play complementary roles in many public rituals. An
example is the council of Rodez convened by Bishop Arnald from his parishes
in 1012, an account of which is incorporated into the Miracula of St Faith.
There the author tells us that it was the customary practice for die bodies
of the saints to be brought to synods, and in accordance with this a veritable
‘battleline of saints’ was drawn up in tents and pavilions, widi golden statue rel-
iquaries of Sts Marius, Amantius and Faidi (die famous reliquary still to be seen
at Conques contains her head), a golden image of the Virgin Mary and a chest
reliquary of St Saturninus. There the common people, the vulgus, also were
gadiered together with lay notables and the clergy; they exclaimed in wonder at
the miracles performed by St Faith. Such councils and religious gatherings
were but one manifestation of the varieties of religious response in Europe in
the tenth and early eleventh centuries. The Peace of God movement in France

Cambridge Histories Online © Cambridge University Press, 2008

The Church 1 6 1

in particular was proclaimed at such councils where, in public gatherings of
clergy and laity and in the presence of the saints, the warriors swore an oath of
peace within society. The popular participation in die Peace of God movement
has been established in recent work, but die episcopal leadership within it and
the bishops’ attempt to exert social control in dieir own terms and by means of
religious fervour is also of crucial importance.

Devotion to saints and investment in prayer is also reflected in the enor-
mous number of donations to the church, particularly to local monasteries,
recorded in the many extant charters of the tenth and eleventh centuries. In
this respect there is clear continuity with die eighth and ninth centuries. The
monasteries are a most positive and visible reflection of lay piety. Monasteries
ministered as effectively to the religious needs of the population as the parish
churches; they provided, in many cases, the essential link with the holy and an
alternative visible contact with God and his saints to that provided by the
secular church. In addition, the charters from such monasteries as Bobbio,
Nonantola, St Gallen, Weissenburg, Lorsch, Saint-Benigne at Dijon or S. Maria
at Ripoll reflect how these houses provided one compelling focus of religious
devotion and loyalties and service to the cult of a local saint. Grants made to
the monastery were insurance against damnation and provided for a donor’s
immortal soul. Yet diey were also a token of the lay Christian’s participation in
the religious life and a very particular contribution to the promotion of die
Christian faith of which every assessment of lay piety needs to take account.
Not only did the laity contribute the material base on which a monastery’s live-
lihood depended. They also contributed their sons and daughters to the service
of the Christian faith. No doubt in some cases pious motives were not the sole
determinant in what prompted die laity to adopt the monastic way of life. The
need for refuge, physical unfitness for a secular life, or social and political pres-
sures no doubt played a role as well. Nevertheless, die astonishing varieties of
monastic life documented by Wollasch (chapter 6), as much as the successful
expansion and consolidation of die secular, episcopal and parochial church
described in this chapter, were essentially a consequence of the laity’s response
to the demands and needs of dieir religion. The building of churches, the
establishment of pilgrimage centres, the devotion accorded relics, the conven-
ing of synods, the concentration, in personal and public morality, on Christian
norms, and participation in the forsaking of comfort and home in order to go
on pilgrimage to die many new sites which were becoming a focus of devotion,
such as Compostella, as well as the familiar goals of Rome and Jerusalem, and
participation in the liturgical rituals of die church were all possible means of
practising Christianity. The developments within the church in Europe in die
tenth and early eleventh centuries were extraordinarily rich, varied and creative.

Yet in all die topics treated in this chapter - ecclesiastical organisation, the

Cambridge Histories Online © Cambridge University Press, 2008

l6z ROSAMOND McKITTERICK

history of the popes, synodal legislation, missions, canon law, liturgy, music,
popular devotion - it has been clear not only how important tire eighth- and
ninth-century Carolingian foundations were but also how crucial the leader-
ship of the bishops became in the course of the tenth century. To some degree
this is what the authors of the many vitae of bishops wished us to think. Yet
their advocacy, and the fact that there is not the same concentration of episco-
pal vitae in the ninth century, is borne out by tire great variety of other kinds of
evidence considered in this chapter. In the acceptance, on the part of many
modern scholars, of the reform rhetoric of die late eleventh- and twelfth-
century bishops and abbots, however, with its stress on the decadence of the
clergy as a whole and on the evils attendant on a church in die power of the
laity, insufficient acknowledgement has hitherto been accorded the great
achievements of their tendi- and early elevendi-century predecessors in the
expansion and consolidation of Christianity in western Europe.

Cambridge Histories Online © Cambridge University Press, 2008

CHAPTER 6

MONASTICISM: THE FIRST WAVE OF
REFORM

Joachim Wollasch

in every medieval century in Europe monastic life was renewed, and
renewal might indeed be said to have been a characteristic of medieval monas-
ticism. Yet although there is a rich literature on Carolingian monastic reform in
the age of Charles the Great and Louis the Pious, for example, it has become
customary to describe the monasticism of the tenth and eleventh centuries as
‘reform monasticism’, just as one talks of the ‘reform papacy’ and of die ‘era of
reform’. Is there not here perhaps an inherent contradiction, just as diere is in
the chapter-title, suggested by the editor of this volume? For if one talks of a
‘first wave of reform’ in connection with the monasticism of die tendi and
eleventh centuries, then this must imply, considering how much we know
about Carolingian monastic reform, that tenth-century reform was a much
deeper caesura in the history of medieval monasticism, a fundamentally new
beginning. Recently Gerd Tellenbach has described the period from the ninth
to the eleventh century as ‘the great era of monasticism’, but he did not talk of
reform monasticism and indeed he questioned the notion of monastic deca-
dence in die tenth century . 1 There were indeed many monasteries ruined by
warfare, and in west Francia they suffered not only from Norse attacks but also
from the long-drawn-out transition from Carolingian to Capetian rule with its
concomitant transfer of power from an increasingly weak centre to local and
regional lordships, a process which afflicted monasteries in particular. It is
probably no accident that it was precisely in Aquitaine, Burgundy and
Lodiaringia that expectations of the coming end of the world emerged
decades before the critical year 1000. Odo of Cluny wrote of the dangerous
times and the threatening end of the world: now die time has come, now
Antichrist stands before the gate. ‘It is precisely this repeated “now” which dis-
tinguishes the tenth from the ninth century .’ 2

Tellenbach also quite rightly criticised die view that it was monastic proprie-
tors who should be blamed for all die damage done to coenobitic monasticism

1 Tellenbach (1988/1993), p. 101. 2 Fried (1989), p. 413.

163

Cambridge Histories Online © Cambridge University Press, 2008

164

JOACHIM WOLLASCH

Cambridge Histories Online © Cambridge University Press, 2008

Monasticism: the first wave of rform 1 6 5

during the tenth century. He pointed to those ecclesiastical and lay proprietors
of monasteries who supported and endowed their monasteries in such a way as
to allow a flourishing of monastic life and so demonstrated their lively interest
in their foundations; we shall return to some of tire more prominent tenth-
century examples later. Moreover, many monasteries, in the German kingdom
in particular, did not decline during die Ottoman era from die level they had
reached in the Carolingian period. There were many monastic reforms, not just
one, and diey need to be examined individually. We can indeed note the impres-
sive forces working for a renewal of monastic life in the tenth century; but in
order to speak of reform monasticism in the tenth and elevendi centuries and
of a movement which experienced a ‘first wave’ in the tenth century, we would
need to find evidence of symptoms common to all the coundess monastic
renewals, in particular to find signs of an awareness within contemporary
monasticism of a need and indeed an imperative to reform monasticism as a
whole and not just one’s own monastery.

To investigate the customs of coenobitic monasticism in die tenth century
no longer entails a wearisome process of gathering together details from the
most varied sources - capitularies, accounts of monastic administration,
letters, petitions, saints’ Lives - as it does for the ninth. By the turn of the mil-
lennium the written codifications of coenobitic monastic custom had become
a separate genre, though with a great variety of titles, a source whose numbers
grew steadily dirough to the twelfth century. The majority of these consuetudines
also include - and in this respect drey did not draw on the Rule of St Benedict
for precedent or inspiration — provisions for the written recording of the com-
memoration of the dead in the office-book of the chapter (in lihro vitae , in lihro
regulae , in martyrologio , and so on). Such chapter office-books become increas-
ingly numerous in the period before and following the millennium, though
many survive only in fragmentary form. Evidendy the commemoration of the
dead was taking on greater importance for monastic communities and for their
sense of diemselves, while at the same time die attractiveness of the monaster-
ies for die laity came to lie more and more in die commemoration of the dead,
which could be performed by die monastic communities from anniversary to
anniversary over generations. In the Tegernsee letter-collection, for example,
we find a letter of thanks from Abbot Gozbert addressed to a special benefac-
tor of the monastery, Count Arnold: ‘your memory has been preserved up to
now in assiduous prayers, but henceforth we have decreed that your name is to
be held memorially fast in our monastery in prayers day and night’. 3 Just how
much such a commemoration was sought after emerges from a further passage
in the letter. For the count’s deceased wife the customary commemoration had

3 . . pro qua extunc usque nunc consuetudinarias complevimus precaminum celebrationes et in

semper annuali revolutione temporis vigilias missasque cum oblationibus sciamus facere’, no. 22 , p. 2 3 .

Cambridge Histories Online © Cambridge University Press, 2008

JOACHIM WOLLASCH

1 66

been practised. Now the community was proposing to celebrate vigils and
Mass with offerings every year. The abbot’s request to the count shows that this
meant yearly commemoration: ‘Let die date be written on a sheet of parch-
ment and send it to us by this present bearer of the roll of the dead .’ 4 What was
now being commemorated in monasteries was no longer the collective
memory of die many people whose names had been collected in the books of
commemoration since die Carolingian period; rather, it was the commemora-
tion of the individual, carried out on the date of their death and renewed yearly
for more and more people.

These codifications of the customs of coenobitic monasticism often
implied a claim that they should be adopted by others, to set out customs of a
model character. A famous example from the tenth century is die Regularis Con-
cordia for die monasteries of England. Here, however, one must distinguish
carefully, depending on whether die model character of such consuetndines wzs
ascribed to the monastery from without - by odier monasteries for example, or
by kings, bishops or abbots - or whether a community itself intended to dis-
seminate its consuetndines among other monasteries. This in turn raises a furdier
point. The spread of two central kinds of text - consuetndines and the chapter
office-book — up to the turn of the millennium allows us to speak of reform
monasticism. But we must distinguish between those cases where die initiative
for reform proceeded from an ecclesiastical or lay monastic proprietor and
those where a reform-minded chapter cooperated with a ruler: was it a royal
order, the expression of episcopal will or the requirements of aristocratic
interests which led to reform, or was it the will of die community of abbot and
monks itself? Such a distinction is lacking, for example, in Kassius Hallinger’s
well-known work Gorge— Kluny? the tide sets the monastery of Gorze, a pro-
prietary monastery owned and reformed by the bishops of Metz, on the same
level as the monastery of Cluny, which carried out reforms on its own initia-
tive, and the tide equates Cluniac monasticism, whose caput Cluny was free of
all spiritual and temporal dominatio, widi that of other ‘reform monasticisms’
practised in monasteries each of which had its own lord and above these a king.
We are thus not dealing with a single reform monasticism including varied
observances and opposed tendencies. Any account of monastic reform in the
tenth and eleventh centuries must allow for the fact drat most monasteries and
monastic groupings were locked into structures of lordship and could not
move outside narrow limits, regardless of whether they were to be reformed by
order from above or whether they wanted to reform diemselves and odiers.
The crucial questions are these. How far in such circumstances was indepen-
dent initiative for monastic reform possible and how did the free monastery of

4 ‘Diem kalendarium iubete conscribi membrana nobisque transmitti per presentem pelligerum’, ibid.

5 Hallinger (1951a, 1951b).

Cambridge Histories Online © Cambridge University Press, 2008

Monasticism: the first wave of reform i 67

Cluny conceive of and carry out reform while absorbing other monasteries
into an ever-growing monastic group? Where was monasticism renewed, and
where did it itself become a force for renewal? Considered in this way, reform
monasticism appears not as a static phenomenon with its own inherent norms
but rather as a dynamic and multi-faceted movement, which had first to create
centres around which it could crystallise.

EREMITIC MONASTICISM

In the face of die well-known monastic foundations of the tenth and eleventh
centuries it has often been all too easily forgotten that there were also eremitic
tendencies in reform monasticism from the beginning of the tenth century
onwards, and that these were more independent of lordly influences. We need
not here diink only of the male and female inclusi who lived in or near monas-
teries at St Gallen, Verdun and elsewhere. Odo, later abbot of Cluny, was
accompanied by a hermit Adhegrinus when he left Saint-Martin of Tours,
where he had been a canon, and became a monk in die monastery of Baume.
But there were also hermits who participated in monastic renewal. Benno, who
was summoned in 927 by King Henry I of Germany from the loneliness of
Meinradszelle (‘Meinrad’s cell’), where he had lived since about 905, to the
bishopric of Metz, paid for this by being caught up in the cross-currents of
Lotharingian politics, blinded and driven from his episcopal city. He returned
to his career as a hermit and, togedier with the provost of Strasbourg,
Eberhard, helped to renew monastic life at Meinradszelle in 934. Itis neverthe-
less characteristic of European monasticism that this eremitical initiative
under Eberhard as its first abbot should have led to the foundation of a monas-
tery whose name, Einsiedeln (‘Hermitage’), reveals something of its begin-
nings but which rose under die patronage of the dukes of Suabia to become
one of the most highly privileged monasteries of die whole Ottoman
kingdom. As such it functioned well into die elevendi century as a source of
renewal for odier monastic communities. When the bishop of Metz reformed
his proprietary monastery of Gorze the archdeacon Einold of Toul became its
first abbot under die new dispensation, and he too had begun his monastic
career as a hermit.

Petrus Damiani, from 1043 prior of the colony of hermits at Fonte Avellana
and later cardinal bishop, has described for us in his Vita Romualdi how the
Ravennatese aristocrat Romuald ‘wished to convert the whole world to
eremiticism’, 6 and how he sought in vain to win Otto III for a monastic life in
the marshes of Pereum near Ravenna. Instead of this the emperor urged

6 ‘to turn mundum in heremum convertere volens’, Petrus Damiani, 1 Vita Romualdi , c. 37, p. 78.

Cambridge Histories Online © Cambridge University Press, 2008

1 68

JOACHIM WOLLASCH

Romuald to reform as its abbot the monastery of S. Apollinare in Classe, where
Romuald had spent diree years as a novice. When Abbot Odilo of Cluny
visited Otto III, seeking the societas of William of Saint-Benigne, and stood
together with bishops and abbots before emperor and pope in S. Apollinare in
Classe itself, the list of participants includes ‘Romualdus abbas et eremita’. 7
After Romuald had failed in this task he threw down his staff of office at the
feet of emperor and archbishop of Ravenna and left die monastery. There is
no doubt that Romuald wanted to be a monk, to win souls for monasticism,
and no doubt also that he saw die strictest form of monasticism as eremitism.
It is uncertain whether he adapted the Rule of St Benedict for his companions;
he had certainly become familiar with it in his period in S. Apollinare, and he
also knew Monte Cassino at first hand. As well as Petrus Damiani, anodier
source from the first half of the elevendi century, the Farfese Liber Tramitis,
tells us that Abbot Hugh of Farfa had undertaken to renew the old institutions
of the fathers and especially of Benedict in Farfa, following Romuald’s
example, and ‘shining as an ornament of monasticism’ like Romuald ‘renewed
the norm of ancient justice in both sexes and in both orders [laymen and
monks]’. 8 He had many pupils, and lived the coenobitic life widi them in differ-
ent places; indeed he founded monasteries himself before dying at Camaldoli,
which in the course of the eleventh century collected a whole group of eremit-
ical colonies around itself. The hermit Nilus, who came from Byzantine south-
ern Italy and so deeply moved Otto III, also acted as abbot and wrote a hymn
to Benedict of Nursia.

The careers of Benno and Eberhard of Einsiedeln, Einold of Gorze and
Romuald all show that the eremitic monasticism of die tenth and eleventh cen-
turies could not remain independent alongside coenobitic monasticism - it
was, after all, not exacdy the same as die way of life of the ancient hermits in
the Egyptian desert - and also diat its representatives found it almost as diffi-
cult as did their coenobitic counterparts to escape the induence of their lord’s
conceptions of monastic renewal and indeed of his demands on their service.
As a partial explanation we may note that they were all of high aristocratic
origins and that many had held high ecclesiastical office before being sum-
moned to hold bishoprics and abbacies. Although these anchorites of the
tenth century, who did not shrink from speaking die blunt truth even to the
emperor’s face, are impressive figures, dieir careers show also the tensions
between ascetic life in the wastes and the stricdy regimented coenobitic life in a
monastery which pulled these eremitical fathers this way and that.

It was coenobitical monasticism rather than eremitism which set the route to

7 DO III 396.

8 Liber tramitis aevi Odilonis abbatis , p. 3 : ‘decore splendidus monachico Romualdus nomine qui normam

priscae iustitiae in sexu renovavit utroque et ordine’.

Cambridge Histories Online © Cambridge University Press, 2008

Monasticism: the first wave of reform

169

the climax and turning point of European monasticism in the twelfth century.
Examples of this are also provided by monasteries like the Reichenau, St
Gallen, Fulda, Hersfeld, Corvey, Lorsch and countless others, all of which
were by and large able to maintain die high level of monasticism achieved in
the Carolingian period through the Ottoman era and beyond. Nevertheless,
this does not mean that they were of continuous importance outside their own
immediate surroundings: their wider intiuence in die Empire and the other
European kingdoms depended on several factors, not least on the priorities of
kings and emperors, which changed from dynasty to dynasty and from ruler to
ruler.

LORDSHIP AND REFORM

If one were to depict the reforms of the tenth century in strict chronological
order of the date of foundations of those monasteries which became focuses
of reform activity, one would have to begin with the foundation of Cluny in
910. However, the testament in which the first abbot of Cluny, Berno, defined
his monastic inheritance dates from 927, when Abbot Odo, under whom
Cluny became a centre of reform, took over the monastery. Odier prominent
centres of monastic reform may thus be considered first, with Brogne in the
diocese of Liege coming before Gorze near Metz. These names remind us also
that both old and new foundations could become sources of renewal, a further
reason for not following die dates of monastic foundations. Any chronology
presents us with the thorny problem of evaluating the various reform initia-
tives. What we find is a simultaneous multiplicity of monastic reforms in the
most varied situations both in old-established and in newly founded monaster-
ies. In the old Carolingian core-land of Lotharingia (Flanders, Lorraine,
Burgundy, upper Italy), aristocratic and royal monastic proprietors competed
with one another in setting up exemplary monastic communities, but there
were also impressive reforms in Rome, southern Italy and England, as well as
in many parts of the east Frankish/ German kingdom. What is common to
them all, with the exception of Cluny, is one tiling: the proprietors’ will to see
monasteries functioning to the highest spiritual standards was coupled with die
willingness of monastic communities to reform themselves and others. The
more we — quite legitimately — look for currents of tradition running from the
Carolingian monastic reforms of die era of Benedict of Aniane through to the
reform centres of the tenth century, the more we risk losing sight of the ele-
mentary connection between lordship and reform, a connection not to be
grasped simply in legal or constitutional terms. It was the proprietary lord and
founding abbot Gerard who placed his monastery of Brogne under the protec-
tion of the bishop of Liege - where it was soon to be granted privileges by the

Cambridge Histories Online © Cambridge University Press, 2008

JOACHIM WOLLASCH

I 7 0

Ottomans - before offering his services to the margraves of Flanders, for
whom he reformed a number of monasteries, not just St Peter in Ghent. In his
capacity as lord of Brogne he served not only Margrave Arnulf I of Flanders
but also Duke Gislebert of Lotharingia, for whom he reformed Saint-Ghislain.
While his Life suggested that he led a sheltered existence ( yitam theoreticani)
within the cloister next to the entrance, the Miracula sancta Gisleni note that he
directed many monasteries at that time. But this direction had been entrusted
to him by the responsible monastic proprietor, occasionally even with refer-
ence to the emperor. 9

In Gorze, the part played by the monks in the reform of the monastery,
which belonged to the bishops of Metz, is more clearly visible than in the case
of Gerard of Brogne, because we have a whole series of names of men who
formed a group of monks leading an eremitical life together with the archdea-
con Einold of Toul already mentioned and with John of Vandieres, who came
from a humble background on the estates of Gorze, of which he was later to
be abbot. After John had made a journey to southern Italy, the whole group -
Bernacer, a deacon from Metz, Salecho of Saint-Martin near Metz, Randicus
of Saint-Symphorian in Metz, and later Frederick of Saint-Hubert in the
Ardennes, Odilo from Verdun, Angelram from Metz, Andrew, Isaac and
the recluse Humbert - was on the point of quitting die woods around Metz for
the soudi. In 933 Bishop Adalbero I of Metz gave them the monastery of
Gorze as a place to live in and renew. Until then Count Adalbert had acted as
abbot, having been granted it by Wigeric, formerly bishop of Metz and abbot
of Gorze. Adalbert now restored die monastery to Bishop Adalbero. A pre-
condition for reform in Gorze was thus that the episcopal proprietor should
commit the task to this group of eremitical monks, and the reform itself had to
be begun by Adalbero, who secured the restoration of the alienated lands of
Gorze, which had been granted out as benefices to die bishop of Metz’s mili-
tary following. One might thus almost talk of a refoundation of the monastery.
The monks mentioned above, who were familiar both with the Rule of St
Benedict and with die views of Benedict of Aniane on monasticism, intended
simply to lead a monastic life.

Up to die end of die tenth century, according to die evidence of a Gorze
necrology which has survived in fragmentary form, fourteen monks of the
monastery had been called to be abbot in the monasteries of Senones (twice),
Stavelot, Saint-Hubert, Saint-Arnulf in Metz, Saint-Michel-en-Thierache,
Saint-Martin near Metz, Saint- Aper in Toul, Moyenmoutier, Saint-Nabor,
Sainte-Marie-aux-Martyrs, Ellwangen, Marmoutier and Saint-Vincent in Metz.
A few examples may serve to show who took the initiative in these promotions.

9 1 /ita Gerardi abbatis Broniensis, c. 1 5, p. 665, and Rainer, Miracula S. Gisleni , c. 10, p. 5 84; cf. Smet (i960),

PP- 44, 5°-

Cambridge Histories Online © Cambridge University Press, 2008

Monasticism: the first wave of reform

D 1

Before the Gorzian monk Odilo of Verdun became abbot of Stavelot, the
monastery had been held by Duke Gislebert of Lotharingia. At Saint- Arnulf in
Metz, where Aristeus (and before him Harbert, not mentioned in the necro-
logy) became abbot, Bishop Adalbero I had replaced the canons by monks
from Gorze. Frederick, an uncle of Adalbero, had left Saint-Hubert because he
could not lead the monastic life there; he became prepositus in Gorze and
returned to Saint-Hubert as abbot after Bishop Richer of Liege, uncle of
Count Adalbert of Metz, had restored coenobitic life there. Senones was also
owned by the bishop of Metz.

It cannot be doubted that die participants were conscious of the important
role played by the bishopric of Metz widiin the royal church of the east
Frankish/German kingdom. In Saint- Arnulf, Metz and Saint- Aper, Toul, as in
Senones, Otto I’s consent was sought for the reform of die monastery.
Marmoutier in Alsace also belonged to die possessions of the bishop of Metz.
In the monasteries which received Gorzian monks as their abbot and Gorzian
monasticism, the ordo Gorfensis, it was the bishops and counts of Metz and
their relatives who dominated. The Gorzian reform in diis way brought
Lodiaringian, and especially Metz monasteries closer together, but it created
neither a monastic order as a legally defined organisation nor a network of
daughter houses joined by a common observance. Nor did Gorze become a
centre of reforming monasticism, even though Einold had a high reputation
within the kingdom and John, the second abbot after the reform, had served
Otto I on the recommendation of the bishop of Metz as an ambassador to the
caliph of Cordoba.

It was die old abbey of St Maximin at Trier which was characteristically used
by the Ottomans when they wanted to renew or found a monastery in their
kingdom. After Abbot Ogo of St Maximin had reformed his monastery with
the help of monks from Gorze and at the suggestion of Duke Gislebert of
Lodiaringia, Otto I promoted him to the bishopric of Liege. It was with monks
from St Maximin that Otto I set up the first community at Saint-Maurice in
Magdeburg, and from there that he summoned Adalbert to be the first arch-
bishop of Magdeburg. Otto’s brother Brun, archbishop of Cologne, sum-
moned a monk from St Maximin to be the first abbot of his foundation at St
Pantaleon; this took place during the pontificate of Ekbert, a relative of the
Ottomans, in Trier, showing once again the importance of family connections
for reform. Ogo of St Maximin had reformed his monastery, which was not
dependent on Gorze, in friendly cooperation with monks from Gorze. When
these once again contemplated leaving Gorze in the somewhat constrained
period following die renewal there, it was Ogo who offered to take them in at St
Maximin. From his time onwards and as a result of Ottoman patronage St
Maximin, intellectually and spiritually orientated towards Gorze, became a

Cambridge Histories Online © Cambridge University Press, 2008

1 7 2

JOACHIM WOLLASCH

centre of reform, influencing the imperial monastery of Weissenburg,
Ellwangen, Echternach, St Gallen, and Gladbach (a proprietary house of the
archbishops of Cologne); Sandrat of St Maximin played an important role
here. Even in tire later tenth century Otto II still used St Maximin monks to
reform Tegernsee, and Bishop Wolfgang of Regensburg was to summon
Ramwold from St Maximin to become abbot at St Emmeram in Regensburg.
Such examples may serve to show that neither Gorze nor St Maximin was a
centre of a self-propelling monastic reform movement establishing daughter
houses which followed a similar observance. What cannot be overlooked is the
link between the will of imperial bishops in respect of their proprietary mon-
asteries or of the Ottomans in respect of the royal monasteries on the one
hand and on the other the readiness of eremitical and ascetically living monks
to live a coenobitic life. The most distinguished abbots became transmitters of
monastic reform in offering their services to rulers.

When we look at the beginning of the eleventh century and compare those
monasteries which under Henry II had replaced the main centres of the
Ottoman era, then the connection between lordship and reform becomes
evident once more. Gorze itself in the second decade of the eleventh century
came to seem so much in need of renewal that Bishop Theoderic II of Metz
summoned the abbot of Saint-Benigne, Dijon, William of Volpiano, to Gorze
as abbot. He had already reformed tire Metz abbey of Saint- Arnulf on behalf
of Theoderic’s predecessor. The reform of the English monasteries should
also not be forgotten in the context of the interplay of lordship and reform. It
is true that recent research by Hanna Vollrath has somewhat modified tire pre-
vious received view, 10 but there can be no doubt of the cooperation of Edgar
and his wife with the abbot-bishops Dunstan and fEthelwold and with other
bishops, abbots and abbesses, which aimed at an ordered monastic life within
the whole of the patria. Dunstan had gone into exile at St Peter’s, Ghent, which
had been reformed for Margrave Arnulf of Flanders by Gerard of Brogne.
After his return, monks from Ghent were summoned to reform English mon-
asteries. But the Regularis concordia, on which those assembled at Winchester c.
970 agreed as a general and binding statement of coenobitic monastic custom,
mentions the monachi Floriacenses even before tire monachi Gandauenses. Contacts
had already been made by Fleury with English monasteries before the death of
Abbot Odo of Cluny in 942, who had reformed the royal monastery. At the
king’s orders, so says the Regularis concordia, tire abbots and abbesses had sum-
moned monks from these two famous monasteries so that they could gather
everything which might serve honourable and regular monastic life, just as
bees gather pollen from flowers, and bring it together in the book of the

10 Vollrath (198;), pp 274-85.

Cambridge Histories Online © Cambridge University Press, 2008

I 73

Monasticism: the first wave of reform

Regularis concordia } 1 If one recalls how strongly monks were bound to the
observance of their own house when they made their profession, then it will
be clear just how much any attempt to impose uniform customs on all monas-
teries in Edgar’s kingdom must have been directed by die royal will. Just as
Benedict of Aniane had laid down una consuetudo for all monasteries on behalf
of Charles the Great and Louis the Pious, so the Regularis concordia wzs a means
of unificatory direction of monastic life, a means about which there was no
dissent between the abbot-bishops and the king. This is confirmed by the pro-
vision that the elections of abbots and abbesses should take place with the
king’s agreement and following the directions of die Rule of St Benedict — the
two conditions are placed in this order by the Regularis concordia\ The customary
singing of psalms for king and benefactors, ‘by whose donations we are nour-
ished through Christ’s gifts’, 12 in the words of the same text, was not to be per-
formed too quickly. The abbots and abbesses of the monasteries should come
to perform obedient service for the king and queen as often as was necessary
for their houses. The powerful were not to use the monasteries for convivia, but
were rather bound to the benefit and to die defence of these houses. As early
as di e. generate concilium held c. 967 Archbishop Dunstan of Canterbury, in accor-
dance with die wishes of Pope John XIII, had threatened clerics that they
would lose their churches should they not observe the obligation of celibacy,
and indeed the king substituted monks for such clerics throughout his
kingdom, especially where they inhabited former monasteries: the secular
cathedral clerics in Canterbury, Worcester and Winchester in particular were
replaced by monks.

Even allowing that King Edgar, much praised in tradition as the founder of
forty monasteries and renewer of many others in England, claimed no more
rights than were at diat time accorded die king in any case, and even assuming
that die abbot-bishops did no more than exercise their inherent powers and
allowed the abbots and abbesses to choose whether diey would agree to such
changes on behalf of dieir houses, we can still not overlook just how near the
prologue to die Regularis concordia stands to that text of 966 written in gold
letters in the manner of a charter, which catalogues the royal acts in favour of
the monasteries, nor ignore the fact that it is precisely die monastic text of the
Regularis concordia which contains formulations showing just how dominant was
the role played by the king in monastic reform. No doubt the abbot-bishops
dEthelwold and Dunstan saw royal support as a conditio sine qua non for success-
ful reform, and hence they took care to depict reform as a process correspond-
ing to die royal will: for this reason monastic reform under King Edgar reveals
a reform monasticism dependent on royal authority: the male houses were,

11 Regularis concordia Anglicae nationis, pp. 72 — 3 . 12 Ibid., p. 74 .

Cambridge Histories Online © Cambridge University Press, 2008

174

JOACHIM WOLLASCH

according to the third chapter of the Regularis concordia, under the protection of
the king, while the nunneries were entrusted to that of the queen.

A final prominent example of the ruler’s initiative in connection with
monastic reforms in the tenth century can be seen in the actions of Alberic II,
princeps atque omnium Romanorum senator , 13 but because this initiative coincided
with another, that of Abbot Odo of Cluny, it will be more convenient to
discuss this example in connection with the development of Cluniac reform
monasticism.

CLUNY

Cluny differed from all die centres of monastic reform previously mentioned,
though there too all the monks wanted to do was to live according to die Rule
of St Benedict as amplified by the tradition of Benedict of Aniane. What has
always been adduced in discussions of Cluny’s unique rise to European signifi-
cance is the freedom from all ecclesiastical and secular lordship granted at its
foundation, and Duke William the Pious of Aquitaine did indeed in 909/ 10
renounce all proprietorial rights for himself and his heirs and place his newly
founded monastery ‘under the protection, not the lordship’ 14 of the pope.
Such a transfer of a monastery to the protection of the Holy See had already
been practised on occasion in the Carolingian era, but within the overarching
order of the Carolingian empire the act had a different meaning from the tradi-
tio of Cluny at the beginning of the tenth century; it has been noted frequently
enough by scholars that die pope was then in no position to protect the monas-
tery offered to him. The new foundation in the valley of the Grosne lay in the
west Frankish kingdom, in a region which from the second half of the ninth
century onwards had not been close to the king; die foundation charter was
dated later than Charles the Simple’s recovery of kingship on the deadi of King
Odo, yet it named King Odo as the feudal lord of the ducal founder and stated
that the foundation was also for the remedy of Odo’s soul. Cluny was thus,
once William had renounced his own claims and those of his heirs, founded
outside royal lordship, indeed outside lordship of any kind; unlike other mon-
asteries, it had no lay or ecclesiastical patron. It was thrown back on its own
resources, and had to fill the vacuum left in order to secure the freedom it had
received. At the time it must have seemed a considerable gamble whether this
would succeed. Sixteen or seventeen years after the foundation Abbot Berno in
his testament described Cluny as a place which was still incomplete after
William’s and his own death: 15 poorly endowed and low in number of brethren.

13 Zimmer mann, Papsturkunden , no. 85, p. 147.

14 D Ra 1 2, p. 51: ‘apostolicae sedi ad tuendum non ad dominandum’.

15 Berno, Testamentum , cols. 85 3—8.

Cambridge Histories Online © Cambridge University Press, 2008

Monasticism: the first wave of rform 1 7 5

If on the other hand one recalls diat as late as the early eleventh century Abbot
Odilo had to have King Robert II of France confirm that no one might erect
fortified towers on the monastery’s lands dais will give some idea of how long
and how hard Cluny was pressed by the nobles of the region.

Although Cluny was so small and defenceless, Abbot Berno - who was also
abbot of Gigny, Baume and an unidentified monasteriumAethicense with a depen-
dency at Saint-Lothain as well as of Massay and Bourg-Dieu - neverdieless
chose it as his place of burial and made gifts to it. Whereas in the first three
houses named above he installed his relative Wido as his successor, he
appointed Odo to succeed him in Cluny, Massay and Bourg-Dieu; the testa-
ment explicitly mentions die consent of the monks to this double designation.
Neverdieless, Odo had to ask the pope to mediate in the dispute between
Cluny and Gigny after Wido had challenged the validity of Berno ’s will, and the
pope in turn commended Cluny to King Radulf of west Francia. Cluny thus
formed a group together with other monasteries from the beginning, thus ena-
bling one member of die group to be supported by the others in crises if nec-
essary. But the abstract notion of a monastic network ( Klosterverhand in
German) says little about the nature of the group. When Ebbo of Deols, a
vassal of Cluny’s founder, himself founded Bourg-Dieu in 917, he gave it the
same freedom as Cluny had and forbade his heirs to disturb the freedom of the
monastery under the cloak of advocacy and protection. As mentioned above,
Berno’s testament shows that the new monastery was commended to the
abbot of Cluny in personal union, while a relative of William the Pious, Count
Bernard in Perigord, transferred his monastery at Sarlat to the power of Abbot
Odo. In 929 Countess Adelaide, abbatissa'm Romainmotier, placed her house in
a state of personal union with Cluny. Sauxillanges was made over permanently
to Cluny by its founder, William the Pious’ nephew, as were Souvigny, Charlieu,
and the monastery of S. Maria on die Aventine in Rome. Monastic proprietors
who acted in this way were evidently convinced that it raised their own rank
when their foundations collectively practised an exemplary monasticism under
the leadership of Abbot Odo of Cluny and were protected from injury
through their own heirs by being made over legally to Cluny. In this fashion
Cluny grew to be the head of a group of monasteries which participated in its
freedom. This was a unique arrangement, and shows how Odo caught Cluny’s
defencelessness in the safety-net of his own connections. For the whole raison
d’etre of this group of monasteries and of the central position held within it by
Cluny lay in die dense and comprehensive work of reform created by Odo with
his monks and assistants.

Particularly characteristic here was the way in which Odo made use of con-
fraternity for the purposes of monastic reform, and in this way brought mon-
asteries of quite differing legal status together in a community of the common

Cambridge Histories Online © Cambridge University Press, 2008

i 7 6

JOACHIM WOLLASCH

monastic life. Fleury had been granted to Count Elisiard by King Radulf;
Elisiard in turn had asked Odo to reform die monastery, and he was in the end
able to overcome the resistance of die monks, who feared for the legal status of
their monastery; he corrected those monks who merely followed the Rule
incorrecdy, and at the same time sent four monks religionis gratia to Saint-
Martial at Limoges, where they were to ask die abbot to grant them confrater-
nity on Odo’s behalf, so that no difference might exist between the monks at
Limoges and Fleury, so that the monks could pass freely between the two
houses, and so that die common way of monastic life might be acknowledged
in all things and die two houses might ‘so to speak form one convent ’. 16 The
four monks then went on to Solignac, where they succeeded in gaining the
consent of abbot and convent to the same agreement. Following this, the con-
tract of confraternity was written down and read out before the convents
before being copied into the chapter office-book for the notice of coming gen-
erations.

To renew coenobitic life Odo needed the consent of monastic proprietors,
and he won over many by his initiative. In the Life by John of Salerno it is said
that ‘he was known to kings, familiar to bishops and dear to the hearts of the
great. The monasteries built in their dioceses they transferred to the ownership
of our father [Odo] so that he might order and improve them according to our
custom ’, 17 an account confirmed by coundess examples. Both lay and eccle-
siastical proprietors asked Odo to bring the forms of coenobitic life practised
in Cluny (the ‘Cluniac way of life’, ordo Cluniacensis) to their monasteries, for
example in Aurillac, Tulle, Sarlat, Saint-Pons de Thomieres, Saint-Marcellin de
Chanteuges, Saint-Benoit de Fleury-sur-Loire, Saint-Julien de Tours, Saint-
Pierre-le-Vif in Sens and various Roman monasteries.

The power of Odo’s reforming initiative may be seen in his sermons. At the
Feast of Peter’s Chair he said, citing Pope Leo the Great (who in turn drew on
the first letter to Peter), that in the unity of faith and baptism all are of royal
race and participants in the office of priesdiood. This was a sentiment evi-
dently well received by the nobles in a land distant from the king. In the first
ever Life written of a holy layman, that of Gerald of Aurillac, Odo offered the
nobility the picture he had formed of the ideal nobleman, who was to use his
power and wealth for die benefit of those entrusted to him by God, so diat
lordship meant justice even for the unfree and service to the poor. The interior
life of such a count was that of a secret monk, with a tonsure under his hat: in
this way even die rich and the powerful might become holy. Such examples
draw attention to a further unique characteristic of the Cluniac reform move-
ment. Odo did not confine himself to bringing his own and other monks back

16 / documents necrologiques de 1’abbaye Saint Pierre de Solignac , p. 557 .

17 John of Salerno, 1 Vita S Odonis abbatis Cluniacensis 11, col. 3 9E.

Cambridge Histories Online © Cambridge University Press, 2008

Monasticism: the first wave of rtform 177

to a strict observance of the Rule. He was also convinced of die need to
reform the whole of Christendom, imbued as he was with a sense that the Last
Days had already begun and the Last Judgement was near. True, the motor for
such reform was to be a renewed monasticism; but Odo had held up, as a
model which was binding not merely on monks, the primitive pentecostal
church recorded in the Acts of the Apostles: ‘This is the mean for monks who
are bound by the social life.’ 18 The foundation charter of Cluny itself says that
the foundation was made ‘for the state and die integrity of the Catholic faith’
and for all Christians, since all are bound together by die one love and the one
faith. 19

We have already cited Odo’s sermon on the Feast of Peter’s Chair: in his eyes
the veneration of Peter was pietas christianae unitatis. Since die renewal of coeno-
bitic monasticism from Cluny was to serve as the basis for a comprehensive
reform, it was only logical that Odo, when he asked for a papal privilege for
Cluny and his monasteries, should not have been content with permission to
take in those monks whose abbots could not provide suitable conditions to live
according to their monastic profession and who had therefore left their mon-
astery in order to improve their conversatio. In the famous privilege of March
93 1 he also secured the provision that the Cluniacs might with the consent of
the owner take a monastery under their lordship to improve it. Odo thus had
the pope provide die Cluniacs with a papal command to carry out their work of
reform by writing a licence into the privilege. Although Cluny could not reckon
with practical protection from the popes, who themselves were hard pressed in
Rome, it was neverdieless helpful for the monastery when the pope wrote at
Odo’s request to the king of France, the archbishop of Lyons, die bishops of
Chalon-sur-Saone and Macon that they should support Cluny. As a reformer in
Fleury Odo also had a privilege issued by the pope which laid down the duty of
the archbishops of Lyons, Tours, Bourges, Sens and Rheims and their suffra-
gans to protect himself and the monastery of Fleury. The more lay and eccle-
siastical magnates could be made to interest themselves for Cluny and its
monasteries - including of course those monastic proprietors who had made
their monasteries over to Cluny - the more certainly Cluny avoided the dangers
of dependence on a single protector and the more independently it could
develop, dianks to a widespread and numerous network of patrons. Cluny as
the head of a group of monasteries which all lay and ecclesiastical office-
holders in die country were obliged to protect: that was the route taken by the
Cluniacs from Odo’s time to their goal of filling the protection-vacuum in
which the monastery had been placed at its foundation.

18 ‘Hie modus est monachis quos ligat vita socialist Odo of Cluny, Occupatio 6, verse 5 83, p. 136.

19 ‘pro statu etiam ac integritate catholicae religionis’, Recited des chartes de I’abbaye de Cluny 1, no. 192, p.

I2 5 -

Cambridge Histories Online © Cambridge University Press, 2008

178 JOACHIM WOLLASCH

We have contradictory information in the sources of the tenth and eleventh
centuries as to who took the initiative in Odo’s reform of monasteries of the
city of Rome and its surroundings. John of Salerno tells us in his Life of Odo
that it was at the urgent request of the pope and that of all the clerical orders of
the Roman church that Odo journeyed to Rome to rebuild S. Paolo fuori le
mura. The destructio monasterii Farfensis, a later source, tells us that Albericus,
Romanorum princeps, was so determined to return his monasteries to the norm of
the Rule that he summoned the holy abbot Odo from Gallia, who at that time
presided over the monastery of Cluny, and had him made archimandrite of all
monasteries around Rome as well as granting Odo his birth-house on the
Aventine to build a monastery there. But Odo had commended himself both
to Pope Leo VII and to the princeps-, he had already secured privileges from
Leo’s predecessor in 927, and among those listed as intervening were Kings
Hugh and Lothar of Italy. When Hugh besieged Alberic in Rome it seemed
appropriate to ask Odo to mediate with Hugh, Alberic’s stepfather, who, fol-
lowing the reconciliation Odo arranged, became his father-in-law as well. We
may well ask whether it was mere chance that Odo was asked to reform pre-
cisely those monasteries which lay on the great arterial roads leading to Rome:
S. Paolo fuori le mura, S. Lorenzo fuori le mura, S. Agnese fuori le mura; cer-
tainly a charter of 941 shows that Kings Hugh and Lothar themselves resided
in S. Agnese. 20 The pope will not have overlooked the fact that Odo cham-
pioned the papacy in its period of weakness not only in his support for die ven-
eration of St Peter but also through his links with Rome. Odo did not make his
first journey only when Cluny ’s interests required it; he had already visited the
city while he was still canon of Saint-Martin in Tours. His conviction that it was
necessary to make a pilgrimage to the prince of die aposdes at Rome is
refiected in his account of the no fewer than seven pilgrimages made by the
count and secret monk, Gerald of Aurillac. The Romanitas of Cluny from its
earliest phase onwards has thus been righdy stressed, and die way in which it
went togedier with the Cluniac will to reform from Odo’s abbatiate onwards
can be seen from Odo’s reforms in the neighbouring monasteries of S. Elia di
Nepi, Farfa, Monte Cassino, S. Andrea in Monte Soracte, Subiaco and Salerno.
He installed Baldwin as abbot in S. Paolo fuori le mura, in S. Maria on the
Aventine and in Monte Cassino; the remaining monasteries were confided to
other pupils. He himself went to S. Pietro in Ciel d’oro in Pavia at the request of
Hugh of Italy.

Although the Cluniac reform attempts in the monasteries of Rome and its
surroundings scarcely affected diese houses permanently - on Odo’s death the
pope asked for monks from Einold of Gorze, who had Italian experience,

20 D. Hugh 57-

Cambridge Histories Online © Cambridge University Press, 2008

Monasticism: the first wave of reform 179

while in Farfa the monks violently expelled their foreign brediren — Cluny’s
links with Rome and the Cluniac presence in Rome and Pavia remained a fact
throughout the tenth century and beyond. After the short intermezzo of the
two Gorzian abbots in S. Paolo the monastery was ruled by Ingenaldus, who
came from the monastery of Saint-Julien of Tours where Odo was buried.
Maiolus followed closely in Odo’s footsteps in the journeys he undertook to
Rome and Italy during his time as abbot (954-94). Through the Burgundian
king’s daughter Adelaide, widow of Lothar of Italy and then wife of Otto I,
Maiolus received a monastery in Pavia which permanendy became Cluny’s
property. Otto is indeed said to have offered him the overall supervision of all
the monasteries subject to him in Italy and Germany, an echo of the continuity
in Cluny’s links with Rome and presence in Italy and a response to the nature of
Cluny itself: unlike so many other monasteries it did not merely radiate reform
influences for a short time but continued as a source of reform well into the
Cistercian era. Cluny’s leadership of a group of monasteries was one reason
for this, but hardly the only one. It has rightly been pointed out that from the
middle of die tenth to the early twelfth century Cluny had die good fortune to
be ruled by only three abbots, all outstanding. This was an essential component
of Cluny’s steady development. Cluny shared the privilege of free abbatial
election with other monasteries, but both its foundation charter and the papal
licence for reform laid down, indeed stressed, that the monks of Cluny were to
choose their candidate without pressure from outside and indeed ‘without
consulting any lord’, 21 and to make him abbot according to the provisions of
the Rule of St Benedict. This provision must have been interpreted very liber-
ally in Cluny, for a custom developed here not foreseen in the Rule. Odo had
been designated abbot of Cluny by his predecessor Berno. Since Berno also
intended that he should take over the monasteries of Deols and Massay it is
conceivable (though not recorded) that the monks in these monasteries had
consented to Berno ’s designation. Since Odo’s successor, Aimard of Cluny (c.
942-f. 9 5 4) appears in Cluniac charters during Odo’s lifetime it is probable that
Odo had designated him successor, especially in view of Odo’s frequent
absences, though once again we are not told whether the convent consented.
What is clearly recorded, however, is that Aimard, ill and blind, made Maiolus
his co-adjutor, and then suggested to the whole convent that Maiolus should
succeed him, since he alone was qualified to take on die office; Maiolus’ con-
sensual election is recorded both in narrative sources and in charters. Maiolus
in turn summoned Odilo to be co-adjutor and designated him as successor
with the convent’s consent. Such a designation of die successor by the ruling
abbot is occasionally found elsewhere, and it was known to the author of the

21 ‘sine cuiuslibet principis consultu’, Pope John XI for Cluny (March 931), Zimmermann,
Papsturkunden , no. 64, p. 107.

Cambridge Histories Online © Cambridge University Press, 2008

JOACHIM WOLLASCH

I 8o

Regula magistri, but in Cluny it was standard practice, though evidently the desig-
nation only acquired legal force after the convent had elected the candidate.

Only when abbot and convent were in agreement could the monastery
remain a centre of reform and the head of a whole group of monasteries, and
only so could it preserve its independence. The cooperation between abbot
and convent was an essential precondition for the attraction exercised by Cluny
on its environment. Particularly revealing among the increasing numbers of
donations made to Cluny by lay and ecclesiastical patrons are those made ad sep-
ulturam , in return for the right of the donor and/or his relatives to be buried in
the cemetery at Cluny and for their memories to be honoured at Cluny beyond
their deaths. The graph of such donations shows a steadily rising curve. Beside
die recruitment of the convent of Cluny from ever more families and an ever-
widening catchment area it is the gifts ad sepulturam - which assured the donors
at their death and thereafter the societas et fraternitas of the Cluniacs and partici-
pation in all the prayers and good works of the monks, ideally the same memo-
rial works as those performed for a monk of Cluny himself — which are the
most reliable indicator of the continual growth in the attraction exercised by
Cluny on its surroundings.

EUROPEAN REFORM MONASTICISM AROUND THE TURN OF THE

MILLENNIUM

We can see how far this evaluation of Cluny as a permanent source of reform
corresponds to historical reality if we take a snapshot of reform monasticism
in Europe around die millennium and compare this with the conditions visible
at the middle of the century. We have already spoken of die uncompromising
eremitism of Romuald and Nilus and the deep impression these made on Otto
111 , and also of the development of Einsiedeln from its eremitical beginnings
to one of the most highly privileged abbeys in the Ottoman kingdom. Around
the turn of the millennium there was not much trace of die impulses given to
reform by Gerard of Brogne in Flanders. Although monks from Gorze and St
Maximin had just been summoned to become abbots in royal and episcopal
monasteries — Romuald by Wolfgang of Regensburg from St Maximin to
Emmeram, Immo of Gorze first to Priim and then to the Reichenau, where he
met with great difficulties - Gorze and Trier themselves then became objects
of reform: Gorze at the hands of Bishop Albero II (who had already had Saint-
Arnulf of Metz reformed from Dijon in 996/7), St Maximin by Henry II
himself. Cluny, by comparison, remained an active provider rather than a
passive recipient of reform.

The monastic centres of gravity within Germany shifted with the change
to the Bavarian line of die Liudolfings: Bavarian monasteries now became

Cambridge Histories Online © Cambridge University Press, 2008

Monasticism: the first wave of reform 1 8 1

prominent alongside the famous Ottoman monasteries, and monks were sum-
moned from St Emmeram to become abbots in Lorsch, Fulda, Corvey and
Bleidenstadt, while Godehard, abbot of Niederaltaich and Tegernsee, went to
Hersfeld before becoming bishop of Hildesheim. At the same time the harsh
treatment of monasteries which seemed to the emperor in need of reform
became clear. The account given by Ekkehard IV - later, but with unusual viv-
idness - of the actions of the Ottomans towards St Gallen makes this evident,
as does that of the Annalista Saxo, who, writing of the rights and coenobitic
customs of the monks of Corvey who opposed the emperor’s desire for
reform, said that they were ‘violently changed by the emperor’. 22 The link
established by Empress Adelaide between die Ottoman court and Abbot
Maiolus of Cluny remained a strong one under Henry II: Henry II even
entered Cluny’s confraternity, where he was entered in the chapter office-book
as an example of an emperor who was worthy of having his memory pre-
served. 23

Not only was Gorze reformed rather than reforming after die turn of die
millennium, by contrast with Cluny: William of Volpiano, the man entrusted
with die reform by Adalbero II, had himself been a professed Cluniac monk
and prior of the Cluniac house at S. Saturnin, whence he had been sent
together widi twelve particularly distinguished Cluniac monks by Maiolus to
reform Saint-Benigne de Dijon at the request of Bishop Brun of Langres (a
relative of William’s). William’s reforming activity was thus Cluniac through
and through, as can be seen in his consuetudines. At the same time there can be no
doubt that it developed a momentum of its own: its density, geographical
extent and subsequent induence made William’s reforming work the most sig-
nificant influence on reform monasticism after the millennium. William was
prepared to reform even in those places where Maiolus and Odilo of Cluny
had refused a royal request out of concern for Cluny’s independence. Brun of
Langres confided all the episcopal proprietary monasteries of his diocese
except Saint-Seine to William to reform. The bishops of Metz and Toul
entrusted Saint- Arnulf of Metz, Gorze and Saint- Aper in Toul to him, as men-
tioned. The duke of Burgundy and die count of Blois gave him Saint- Vivant-
de-Vergy and Saint-Faron in Meaux; the king himself, Robert the Pious,
persuaded him to reform Saint-Germain-des-Pres. A further concentration of
Wilhemine reform was found in Normandy, after the duke had summoned
him there, in Fecamp, Jumieges and Mont-Saint-Michel. William founded
Fruttuaria near Turin in 1015/16 together with his own (biological) brothers.
Saint-Benigne’s necrological tradition shows not only that William and his

22 ‘potestative ab imperatore mutantur’: Annalista Saxo, s.a. ioi 5, p. 668.

23 ‘Pro imperatore qui dignus fuerit ita scribatur: Tertio idus iulii depositio domni Heinrici imperatoris

augusti nostrae societatis et fraternitatis karissimi’ Liber tramitis aevi Odilonis abbatis, p. 285.

Cambridge Histories Online © Cambridge University Press, 2008

182

JOACHIM WOLLASCH

monks had strong links with Cluny but also that after die time of William and
his Cluniac companions in Dijon die community there developed its own self-
awareness: the monks of Cluny were no longer recorded among the monks of
Saint-Benigne, but on the facing page of the necrology, reserved for the monks
of die houses who had confraternity with Saint-Benigne. The varying intensity
of the contacts between Dijon and diese other houses is also confirmed by the
necrological evidence. As late as the second half of the eleventh century, con-
temporaries perceived die houses of Gorze and Fruttuaria - which itself
acquired dependencies and whose transalpini monachiwete requested to reform
monasteries as far as the nordi-west of the empire - together with Cluny as the
monasteries of choice for episcopal and lay proprietors who wished to ask for
monks to come to reform their houses.

Saint- Vanne in Verdun under its abbot, Richard, was a further monastic
centre. He wanted to become a monk in Cluny, but had been persuaded by
Odilo that he could do more good as a monk by staying where he had come
from. It is worth noting, incidentally, that the most prominent centres of
monastic reform now lay on die western border of die German kingdom or
beyond it and in upper Italy. We have already noted that Henry II’s reign saw a
shift of emphasis within the monasticism of die German kingdom compared
widi that of the Ottoman era; to this it should be added that the emperor
himself, following a key experience at Monte Cassino in 1022 when St Benedict
cured him of a severe attack of kidney-stones, turned his attention to the
reforming monasticism of the west. Alongside his confraternity with Cluny
there should be mentioned his close links with Saint-Benigne, die rich dona-
tions he made to Saint- Vanne and his concern for the reform of monasteries in
the German kingdom and — together with King Robert II, with whom he had a
meeting at Ivois in 1023 - in France. Fruttuaria also accepted Henry II into
confraternity. The monasteries to which he now gave most attention devel-
oped a greater degree of independence and reform initiative than the monas-
teries which had been centres of reform in southern Germany earlier in his
reign, even if Cluny was here still in a quite different category. Henry sought an
abbot from die west who could carry out a general reform of the monasteries
in his kingdom. It could not be expected of eidier Odilo of Cluny or William
of Saint-Benigne or Richard of Saint- Vanne, all of whom die emperor
revered, diat they would accept a summons to come and assist him. But Poppo
of Stavelot, who in the end yielded to Henry II’s request after he had been
offered the abbey of Stavelot, may count as a pupil of Richard of Saint- Vanne;
he became abbot of St Maximin in Trier in 1022, and then, under Conrad II,
head of the imperial monasteries of Limburg on the Haardt, Echternach,
Saint-Ghislain, Hersfeld, Weissenburg, St Gallen and a whole series of aristo-
cratic and episcopal monasteries.

Cambridge Histories Online © Cambridge University Press, 2008

Monasticism: the first wave of reform 183

The opening up of the royal monasteries of Germany to the reform monas-
ticism of the west brought about by Henry II and by some of his bishops
(Adalbero II of Metz, Gerard I of Cambrai, Meinwerk of Paderborn) was to
take on a new dimension in the era of die Empress Agnes after the middle of
the century. But already after the millennium the signs were increasing that
Cluny under Abbot Odilo now held the key position within European monas-
ticism. Hundreds, of monasteries already belonged legally to Cluny, and more
and more monasteries were from the start founded as Cluniac priories. Under
Maiolus and Odilo the second monastic church was built at Cluny; under Hugh
the third church, the greatest in the west, was begun, and the infirmary at
Cluny, extended under Odilo, became the largest hospital in Europe in Hugh’s
time. As already mentioned, the number of monks had climbed from around
twelve at the beginning to around 1 00 under Maiolus and Odilo; by Hugh’s
death in 1109 it had reached 300 to 400. Even in Odilo ’s time Cluny’s catch-
ment area may be seen as Europe-wide, though western and southern Europe
dominated. What monks did everywhere - prayer for the dead and feeding of
the poor for the preservation of their memory and of their souls — was done in
Cluny more intensively and on a broader scale. Odilo and his convent intro-
duced the celebration of All Souls on 2 November to following the celebration
of All Saints on 1 November, so as to do more than had been customary for
the dead brethren; the feast was subsequently taken up by the Roman church.
In Cluny and its monasteries, not only were the eighteen poor who were to be
catered for every day fed on this day, not only were masses said and offices
prayed for die soul of every dead monk for thirty days following his death, not
only was the daily ration which was ‘his right’ 24 given to a poor person by the
almoner; all poor persons ‘who chanced to pass by’, 25 and whose numbers
could therefore not be calculated in advance, were also fed. The Cluniac consue-
tudines show that from the time of Maiolus the word eleemosynarius refers to a
monastic office-holder responsible for the care of the poor, and the eleventh-
century consuetudines show very clearly how rapidly the duties of die eleemosynar-
ius grew, so that he came to need more and more assistants. The way in which
the care of the dead and of the poor were linked in Cluny under Odilo became
so famous that a whole series of stories and legends grew up around it: that at
the entreaties of the Cluniacs even the pope was granted ease among the pur-
ifying fires; diat nowhere in Europe were so many souls snatched from the
grasp of demons as in Cluny; that Odilo before his death had a monk count up
and write down how many masses he had celebrated during his lifetime. The
Cluniacs were asked whether it was permissible to sacrifice to God daily for the
dead; the answer was that only on Sundays - because of the uniqueness of

24 ‘iustitiam vini et panis’, ibid., p. 265. 25 'omnibus supervenientibus pauperibus’, ibid., p. 1 99.

Cambridge Histories Online © Cambridge University Press, 2008

184

JOACHIM WOLLASCH

Christ’s resurrection from the dead - was it not permissible. The surviving
necrologies from Cluniac monasteries show the intensive and extensive nature
of the liturgical commemoration of the dead at Cluny. They still include the
names of die monks of Cluny in the tendi century — diough by no means all —
and in their unusual numbers and agreement in their entries impressively
confirm what was related in stories and legends.

The community of the living and the dead which thus arose created inter-
nally a sense of Cluniac togetherness; externally it enabled the attraction of
Cluniac monasticism to penetrate its aristocratic, peasant and clerical sur-
roundings, and touch people at all levels of society. Already in 1006 Robert II
could praise Cluny in a charter for Fecamp as ‘the most famous monastery’,
‘whence the source of holy monastic religion has been led in streams to many
places far and wide’. 26 At Romuald’s suggestion Abbot Hugh ‘imposed’ on his
monastery of Farfa ‘die customs of coenobitic life of die monastery of Cluny,
which was built in Gaul and flowered above all other monasteries of its time
over the whole globe on the path of Regular life’. 27 Even before the question of
canonical elections became acute in the course of the so-called Investiture
Contest, Cluny’s self-determination in this respect became as much of a stan-
dard to be matched as was its reputation for exemplary monasticism. In
Robert’s charter for Fecamp it was laid down in respect of the election, installa-
tion and consecration of the abbot diat that custom was to be followed which
had been preserved in Cluny up until then, in other words die absence of any
outside influence. And even before the well-known provisions of die peace
councils were issued, Maiolus and Odilo secured from the bishops assembled
in Anse in 994 a privilege protecting churches, people and property within the
monastery and settlement of Cluny itself, from any act of violence; it also
covered certain named centres of Cluniac property and above all it covered all
monasteries belonging to Cluny. In this privilege, which was matched in
content by a royal charter imprecated by Odilo, general provisions were also
formulated on Nicolaitan priests, on the observance of Sunday, on the duty of
laymen to fast on Wednesdays and Fridays, and on the care of the poor, all
topics familiar from the history of the Peace of God movement. Repentant
infringers of these provisions were to be absolved by the abbot and monks of
Cluny. Even the negative criticisms of the Cluniacs, as offered for example with
satirical exaggeration by Bishop Adalbero of Laon, may serve to show the
influence of Cluniac monasticism on public life. In the bishop’s eyes the
Cluniacs had mixed up the divinely ordained tripartite division of society -
clergy (oratores), warriors ( bellatores ) and the agricultural population (laboratores)
- by allowing peasants to become bishops and bishops to become monks, and

26 DR0II26. 27 Ubertramitis aevi Odilonis abbatis,p. 4.

Cambridge Histories Online © Cambridge University Press, 2008

Monasticism: the first wave of reform 1 8 5

in the way in which Abbot Odilo appeared as rex Oydelo and die Cluniac monk
as miks} %

The preconditions for Cluny’s rise to become the greatest monastery in the
west under Abbot Hugh, who at Canossa was able to mediate between Pope
Gregory VII and King Henry IV, had thus been created. After the middle of
the century a second wave of reform monasticism arose, and Cluny and
Fruttuaria were joined as centres by new ones like Vallombrosa, Saint- Victor in
Marseilles, Hirsau, Saint-Blasien and Siegburg, to name only these few.
Characteristic of diis second wave was its powerful impression on the lay
world, and the monastic call to follow the apostolic life in poverty and in wan-
dering after the example of the ‘primitive church’. It was this second wave of
reform monasticism which culminated in the age of monastic orders, begin-
ning with the Cistercians at the end of the elevendi century. This virtually elim-
inated die risks for the individual monastery within an order, and monastic
self-determination was guaranteed by the order’s constitution and by the
general chapter. The mutual exclusion of the orders, which did not confine
itself to self-definition and caused die older representatives of Benedictine
monasticism to take on the form of an order, also led in the wake of the
ensuing disputes to monasticism’s losing a certain degree of credibility, visible
not least in the appearance of the first heretical movements of European
history in die twelfth and diirteenth centuries. In this period Europe changed
from a landscape of monasteries to a landscape of towns; it was these which
took over from monasteries as crystallisation points of social life. The forces
which concerned diemselves critically with the monastic orders and helped to
secure the defeat of the first heretical movements no longer built great monas-
teries in open countryside as previously, but instead took up residence in die
towns. It was the exponents of Franciscan poverty and die Dominican com-
munity of priests which in turn took on the form of an order as the mendi-
cants.

28 Adalbero, Cawien ad Robertum regem , rses 80—169.

Cambridge Histories Online © Cambridge University Press, 2008

CHAPTER 7

INTELLECTUAL LIFE

Claudio Leonardi

intellectual life in this period is often given labels which relate to other
politico-cultural events and phenomena: die ‘post-Carolingian’ or ‘pre-
Gregorian’ age. The former view clearly conceives of the tenth century as a
continuation of the Carolingian renaissance; by implication, continuation leads
to decadence and finally to the Ottoman renaissance. The latter term sees the
intellectual and spiritual movement of the monastic tradition (in Gorze and
Cluny) and the episcopal tradition (as in Radier of Verona) as precursors of
Gregorian church reform and of the Investiture Contest . 1 Such descriptions
have long left the tenth century without a name of its own, except perhaps for
one of the negative descriptions applied since Baronius’ time: the ‘iron
century’, the ‘dark century ’. 2 Recent reactions against diis have led to the
period’s being described as a great era of cultural renewal and renaissance.
Rather than considering the age as particularly obscure or particularly enlight-
ened, it is more useful to look at what was happening in intellectual life at the
time. The first tiling to consider is schools and book production, which are, at
least in part, closely related.

BOOKS, SCHOOL AND INTELLECTUALS

Fewer manuscripts can be dated to the tenth and the early eleventh centuries
than to the ninth century and the later eleventh ; 3 an oft-repeated fact which
tells us nothing about intellectual life during the century. The boom in writing
from the last years of the eighth century was a result of the huge cultural devel-
opments brought about by Charlemagne. The Carolingian renaissance largely
ended Germanic oral tradition and popular culture, and created a need for a
written culture based on manuscripts. But given the high cost of books, contin-
ued production was unnecessary once demand had been met. This, rather than

1 See in general Secolo di Ferro (1991); Lateinische Kultur (199 1) ; Jacobsen (1985); Hoffmann (1986).

2 Baronius, Annales ecclesiastic, i, col 741 . 3 Bozzolo and Ornato (1983), pp. 1 5— 1 21 .

186

Cambridge Histories Online © Cambridge University Press, 2008

Intellectual life

187

cultural decline, explains die tenth-century fall in manuscript production , 4
though there are some peripheral areas like Benevento and lesser Lombardy
where there are more tenth- than nindi-century manuscripts.’’ For the produc-
tion of written texts to rise, writing must first become a more widespread prac-
tice. The numerous surviving charters clearly show that this was true in legal
contexts; the tenth century can be accurately described as a period in which the
production of documents enjoyed an unusual expansion . 6

Studies of the surviving manuscripts of classical authors, the best examined
field to date, also offer interesting insights into diis problem. According to
Ludwig Taube’s famous remark, the tenth century is an aetas Horatiana by con-
trast with the Carolingian aetas X 'ergiliana ; 7 but Horace was not the only or the
most important new author of the time. The Carolingians had preferred late
classical Christian writers like Sedulius or Arator, as well as Virgil and the
Disticha Catonis , 8 In the late ninth century Remigius of Auxerre, a typical repre-
sentative of the period, was also commenting on Juvenal and Persius, and in an
early tenth-century Milanese codex (Paris, BN lat. 7900A), Martianus Capella’s
De nuptiis Philologiae etMercurii, a good representative of the texts studied in the
Carolingian schools, appeared alongside new authors such as Terence, Horace,
Lucan and even Juvenal . 9 Significantly, the codex is Milanese: Italy had been
less influenced by the Carolingian school and was hence more ready to accept
the work of authors who did not fit within die rigid scholastic codes of the
past . 10

During die tenth century the entry of other classical authors into the scho-
lastic canon became easier and smoodier. But there was also a change in educa-
tional dieory and thus in the tools used by teachers. The curriculum still mainly
consisted of the study of grammar, conceived of as the study of literature
through poetry. Teaching followed a manual, the most important being drat
written in the fourth century by Donatus, wh ose Ars minor h ad been glossed by
Remigius of Auxerre in a successful commentary. During the tenth century
Donatus began to be accompanied, though not yet replaced, by the Institutiones
grammaticae of Priscian as documented in Guadbert’s Epitoma Prisciani. Other
grammars dear to the Irish and Carolingian traditions, such as those of
Pompeus, Consentius, Charisius and Diomedes, also began to fall into disuse . 11
Guadbert realised the continuing importance of Donatus and therefore used
only the first sixteen books of Priscian (the Priscianus maior), and did not adven-

4 Munk Olsen (1991b), pp. 342— 3.

5 V. Brown in Lowe (1980); Brown (1988); Cavallo (1991), pp. 760—1.

6 Toubert (1973a), pp. 303— 5. 7 Traube (i9ii),p. 1 13.

8 Munk Olsen (1991b), p. 345; (1991a), pp. 28-32.

9 Munk Olsen (1991a), pp. 26—7 ; cf. Leonardi (i960), pp. 435—6.

10 Ferrari (1991), p. 108. See also more generally Bischoff (1984), pp. 171—94.

11 Holtz (1991).

187

Cambridge Histories Online © Cambridge University Press, 2008

CLAUDIO LEONARDI

ture into the ‘revolutionary’ territory of books 17 and 18. The importance of
the Epitoma Prisciani was clear: pedagogy was placed at a cross-roads, turning
away from the heavy Carolingian overtones of its own French background and
moving towards the Italian teaching tradition which it discusses. Guadbert was
not an isolated figure; comments and glosses on Priscian were also produced
by Israel the Grammarian 12 and Fromund of Tegernsee. 13

The canon was not simply replaced by a new and equally inflexible system
but altered to fit new needs. The study of grammar lost its Carolingian domi-
nance, being joined by rhetoric (with texts from Cicero and the pseudo-
Ciceronians) and dialectic (using material by Boethius Logicus). 14 At the end of
die century the exceptional but not unique school of Abbo of Fleury
(940- 1 004) 15 and Gerbert of Aurillac (940/ 50-1003) 16 slowly added die quadri-
vium to the three subjects of die trivium and the study of the sciences was added
to the study of literature. The Bible remained the supreme fount of knowledge
and die focus of study for all die liberal arts, but the classical poets had largely
lost their demonic associations: Jerome’s view of the contrast between Christ
and Cicero was now a thing of the past.

The schools were changing internally, but the main innovations took place in
the face diey presented to society. During the Carolingian period schools and
intellectual life ran on parallel paths, and schools were equated with culture;
even imperial culture under Charlemagne was conceived of as a school.
Between die beginning of die tenth century and the beginning of the eleventh,
change in this fundamental identity became ever more evident. Schools ceased
to be totally identified with culture and began to take on a preparatory and
introductory role which continued through the eleventh and the beginning of
the twelfth centuries. Culture, in the strict sense of die word, may have origi-
nated in the schools but it lost its close relationship with them; and it was no
longer in schools or in school- teaching that culture found its dynamic and crea-
tivity.

Typical of this situation were the anti-scholastic debates which seemed to go
beyond the level of topos or habit. Rather of Verona, who had been trained in
Flanders where great schools such as die episcopal school of Liege existed,
had no hesitation in saying of himself, ‘he learned litde from his teachers, much
more from himself’. 17 This was not an unusual attitude to take; rather, it was
symptomatic of many intellectuals of die time, from Liudprand of Cremona
to Gerbert of Aurillac. The changes taking place also meant that men of
culture began to feel the need for a library of dieir own, libraries that would

12 Jeudy (1977). 13 Sporbeck (1991), pp. 369— 78.

14 Van deVyver (1929), pp. 425-52; Gibson (1991). 15 Mostert (1987).

16 Gasc (1986); Riche (1991); Lindgren (1991), pp. 291—303.

17 ‘pauca a magistris, plura per se magis didicit’: Rather of Verona, Phretiesis 11, 70— 1 , p. 200.

Cambridge Histories Online © Cambridge University Press, 2008

Intellectual life

189

exist alongside those of monasteries and bishoprics and alongside the great
court libraries such as that of the new Ottoman emperors from Saxony. 18
Manuscript collecting, characteristic of Italian humanists in the Trecento and
Quattrocento, was also typical in this period, although the movement of learn-
ing was here die reverse, from the north into Italy. 19 This is true of Rather, 20
and also of Gerbert, who wrote to Raynardus of Bobbio, “You know with what
care I seek copies of books everywhere. You know how many scribes can be
found all over Italy in the towns and in die countryside.’ 21

Another important facet of the connection between intellectuals and
schools was the writer’s relationship to his text. The tradition that the author
dictated, or had his work dictated, to a lay or professional scribe came to an
end. Henceforth the author no longer merely checked die work of the scribe,
which the head of the scriptorium continued to do, but became more directly
involved in the text itself, writing or rewriting parts of it, making corrections
and additions. In the 870s Anastasius the Librarian (800-79) had edited his
own translation from the Greek of the Acts of the Fourth Council at
Constantinople, adding extra materials, 22 but this was an exceptional case: since
very few knew Greek, the author alone was in a position to revise his transla-
tion. Almost a century and a half lies between Anastasius and Radulf Glaber (c.
990-1047), during which time die author’s attitude to his text changed consid-
erably: the manuscript Paris, BN lat. 109 12, 23 shows how Radulf wrote and
rewrote his text in different ways. Methods of composition and the intellectual
traditions of this period show that, during the tenth century, cultural activity
continued to take place in the schools and in courts, bishoprics and monaster-
ies, but also diat there was a sense diat only individual study, of manuscripts or
of various fields of learning, really counted. 24

Abelard and the other masters of the twelfth century have been seen as the
precursors of the modern intellectual in that they were connected widi and
taught in a school. They were intellectuals whose job was ‘thinking and passing
on their ideas through teaching’, a job which ‘allied the development of ideas
with the spread of ideas through teaching’. 25 It could be argued, however, diat
the forerunner of the modern intellectual was the tenth-century author; only
during this century did knowledge come to be understood both as being a form
of personal achievement and as having a social role. People were aware of the
dialectical tension between the scholar’s possessive attitude to his subject and

18 Miitherich (1986); Mayr-Harting (1991a and b); McKitterick (1993).

19 Ferrari (1991). 20 E.g. Daniel (1973).

21 ‘Nosti, quanto studio librorum exemplaria undique conquiram Nosti, quot scriptores in urbibus ac in
agris Italiae passim habeantur’, Gerbert of Aurillac, ep. 130, pp. 1 57—8.

22 Leonardi (1967). 23 Garand (1983). 24 Chiesa and Pinelli (1994); Chiesa (1994).

25 Le Goff (1957), p. 1.

Cambridge Histories Online © Cambridge University Press, 2008

190

CLAUDIO LEONARDI

his intensely critical approach to it on the one hand, and the need to demon-
strate publicly (both inside and outside the schools) the knowledge he had
acquired on the other: Liudprand of Cremona, trained in Italy and active also
in Germany at the Ottoman court, was one example of this new kind of
scholar. 26

THE PERIPHERY OF EUROPE

The dissociation, even alienation, of intellectuals from the schools did not
prevent cultural life from being concentrated in particular areas and cities: the
historical and geographical context remains fundamental for our understand-
ing of intellectual life in this period.

Spain, largely occupied by the Arabs, enjoyed an extraordinary blossoming
of Arab art and culture; the developments in Latin literature were insignificant
by comparison. Even taking into account the north of Spain, which had
remained outside direct Arab control, Manuel Diaz y Diaz is right in saying that
few texts showed any conscious literary content. 27 The Latin tradition contin-
ued in Cordoba, the capital of the caliphate during the tenth century. Towards
the end of the ninth century Samson maintained die tradition as did Ciprianus
and Leovigild later; the Cordoba Penitential was perhaps written here, as was a
passion of the St Pelagius martyred in 925 (BHL 6617).

The output of Asturias and Leon was not much richer. Under Alfonso III
(866-9 1 2 )> however, the court of Leon probably did produce the complex text
known as the Chronicle of Alfonso III, diis includes the Chronica Visegotorum , the
Chronicon Alhedense and the Prophetic Chronicle , 28 In Navarre under Sancho
Garces I (906-26), the monastery of S. Millan de la Cogolla flourished again; in
924 the monastery of S. Martino was founded at Albeda, in which Vigilianus
and Sarracinus composed a set of verses. Little else was written, though
towards the end of the century Lupito, who probably lived in Barcelona, pro-
duced a number of scientific texts referred to by Gerbert in his work on the
astrolabe. 29

The situation in England, although very different, did have some features in
common. As in Spain, intellectual life centred on royal courts and monasteries.
King Alfred (871-99) had himself translated several Latin texts he deemed of
value into the vernacular: Augustine, Orosius, Boethius, Gregory the Great,
Bede’s Ecclesiastical History. His translations were part of a plan to raise the cul-
tural and social level of die Anglo-Saxon language, and the tenth century did
indeed see a substantial rise in the production of vernacular manuscripts. But

26 Vinay (1978a); Sutherland (1988); Staubach (1991). 27 Diaz y Diaz (1991), p. 95.

28 Chroniques Asturiennes, ed. Bonnaz and ed. Gil Fernandez; cf. Diaz y Diaz (1981); Lopez Pereira

(1991). 29 Gerbert of Aurillac , Liber de astrolabio.

Cambridge Histories Online © Cambridge University Press, 2008

Intellectual life 1 9 1

this cultural flowering could not have occurred outside the Latin tradition and
it also acted as a stimulus to Latin authors. 30

The first significant work in Latin for fifty years was the biography of
King Alfred by Asser (around 893) but nothing important was then pro-
duced until the reigns of dEthelstan (924-39) and especially of Edgar
(959-75). In dEthelstan’s reign there was a considerable growth in scribal
activity, thanks mainly to increased document production, the import of
manuscripts from the continent, 31 and the arrival of non-Anglo-Saxon
teachers such as Israel the Grammarian. 32 In the years that followed, conti-
nental influence increased via links with Saint-Bertin, Ghent and especially
Fleury. Frithegod worked in Canterbury under Archbishop Odo (941-58)
on his Breuiloquium vitae beati Wilfredi? z Two scholars of outstanding worth,
Dunstan, who worked at Glastonbury, 34 and fEthelwold, who lived at court,
but retired to Glastonbury on Edgar’s death, 35 advanced to high office under
Edgar: Dunstan became archbishop of Canterbury, fEthelwold abbot of
Winchester (where Wulfstan later worked). Oswald, Odo’s nephew, who had
studied at Fleury, became bishop of Worcester and later archbishop of
York; he also founded Ramsey, where Abbo of Fleury taught for a short
time. 36

The Anglo-Saxon court supported a type of monastic reform which
brought with it cultural activity of some distinction. Unlike Spain, which was
rather isolated, Anglo-Saxon culture was influenced by tire monastic reforms
of continental Europe (led by Cluny and Gorze) as well as by the activities of
Europe’s flourishing centres of manuscript production/’ 7 Anglo-Saxon text
production was mainly of hagiographies, generally though not invariably in
verse: lives, translations, miracles (of Swithin, Edmund, fEthelwold and others)
and also fEthelweard’s Chronicon, a Latin translation of a recension of the
Anglo-Saxon Chronicle , 38 At the end of the century fEthelwold’s pupil fElfric,
who became abbot of Eynsham (95 5-1020), represented die highest pinnacle
of Benedictine reform and Anglo-Saxon literature (his works include saints’
lives and homilies). fElfric also wrote in Latin prose and produced a number of
important Latin saints’ lives. 39

In Spain and England tire tenth century was thus a period not of decadence
but rather of renewed interest in study and intellectual activities. Following
the Carolingian model, intellectual life revolved around a royal court and was
supported by the activities of the Benedictine monasteries, a sign of a certain

30 Lapidge (1991a). 31 Keynes (1985). 32 Lapidge (1992).

33 Frithegod, Breuiloquium vitae beati Wilfredi , ed. Campbell; cf. Lapidge (1988).

34 Ramsay etal. (1992). 35 Yorke(i9 8 8). 36 Stafford (1989), pp. 187— 91.

37 Knowles (1963), pp. 31— 56. 38 Lapidge (1991b) Cf. also iEthelweard, Chronicon, ed. Campbell.

39 Gatch (1977); Lapidge (1991a).

Cambridge Histories Online © Cambridge University Press, 2008

192 CLAUDIO LEONARDI

cultural lag on the periphery. It is impossible, however, to categorise the south-
ern and eastern borders of Europe in diis way.

In 88 1 Monte Cassino was destroyed and the monks who fled to Teano and
then Capua were unable to regain the intellectual supremacy they had enjoyed
since the end of the eighth century when Paul the Deacon had been one of
their number. 40 But by the middle of the tenth century the abbey most closely
connected with St Benedict had been completely rebuilt and intellectual life
there had begun again. At the end of the ninth century in Capua Erchempert
wrote his Historic! LangobardorumBeneventanorum. In the second half of the tenth
century in Salerno, an anonymous author wrote the Chronicon Salernitanum? x
and later Laurence of Cassino, who became archbishop of Amalfi in 1030,
wrote sermons and hagiographies. 42

During this period the Latin-speaking scholarly world — most of southern
Italy still spoke Greek, while Sicily was now occupied by Muslims - looked
mainly to Naples and the unusual form of literary activity carried out there:
translations from Greek into Latin. This work had started in the ninth century
and die Neapolitan school continued until well into die second half of the
tenth century. Its aim was to produce a Latin hagiographic corpus for southern
European Christians incorporating existing religious traditions on the saints; as
few of these existed in Latin, works were translated from the Greek. The
school was innovatory: its members moved from die traditional theory of
word for word towards sense for sense translation. Source texts might in con-
sequence sometimes be partially rewritten to meet contemporary require-
ments, but the texts produced were generally of a literary quality far superior to
that of any literal translation. There were numerous author-translators: John
the Deacon, Guarimbotus, Peter the Subdeacon, Bonitus the Subdeacon and
Cicinnius, all of whom wrote many saints’ lives. John of Amalfi, a monk, trans-
lated other hagiographies in the later tenth century; the first Latin translation of
a narrative that would later become very popular in the west, the story of
Barlaam and Jehosaphat, also apparently comes from the Amalfi area. In
Naples at this time Archpriest Leo translated die Nativitas et victoria Alexandri
magni regis of Pseudo-Callisthenes, the basis of another important western
series of narratives. 43

Although southern Italy was on the edges of the Latin world, it played an
important role both by introducing certain aspects of Greek culture into Latin
culture and by confirming the vital role played by historiography and hagiogra-

40 Leonardi (1987).

41 Erchempert, Historic 7 Langobardorum Beneventanorum , ed. Waitz; Westerbergh (1957); cf. Taviani-

Carozzi (1991). For the Chronicon Salernitanum see below at n. 103.

42 Laurentius monachus Casinensis archiepiscopus Amalfitanus opera , ed. Newton.

43 Chiesa (1991); Kratz (1991).

Cambridge Histories Online © Cambridge University Press, 2008

Intellectual life 193

phy in the artistic self-expression of the time, as well as transmitting the narra-
tive, fantastic themes of Barlaam-Jehosaphat and Alexander into western liter-
ature. These are two historical legends of enormous fascination: the story of
Alexander the Great and his Asian empire and the legend of Barlaam, the
Christianised Buddha. The narratives about ancient Rome and the Trojan War
were already popular in tire west but were felt to be linked with the roots of
western civilisation. The new themes introduced by the writers in Naples and
Amalfi represent the arrival of eastern themes in Latin culture.

The eastern borders of Europe were not yet part of the Latin world; these
were missionary lands, dominated by the politico-religious interests of the
Saxon rulers and the pope. It was not until the eleventh and twelfth centuries
that Latin began to be used extensively here. One of the first texts produced
was tire Deliberatio supra hymnum trium puerorum by Gerard, an Italian martyred in
1046 and the first bishop of Csanad in Hungary. 44 The most important texts
related to these countries are the two lives of Adalbert, bishop of Prague, mar-
tyred in Prussia in 997. The first life was probably written by John Canaparius
(who died in 1004) but it has also been attributed to Radim, Adalbert’s brother
and to Gerbert of Aurillac. John Canaparius was monk and later abbot in the
monastery on the Aventine in Rome where Adalbert had spent some time. His
work revives several of the themes of ancient martyr literature, such as testi-
mony to die faith in the face of political oppression; but these themes were
now integrated into the missionary ones typical of die hagiography of the early
and high middle ages. 45

The second life of Adalbert, written by Brun of Querfurt (974-1009), is of
even greater importance. Brun was a chaplain at the court of Otto III but
abandoned it to become a monk in die monastery on the Aventine. He later left
the monastery for a hermit’s life in Romagna as a disciple of Romuald of
Ravenna. His desire for monastic perfection became fused with missionary
zeal and eventually Brun followed in Adalbert’s footsteps. In 1 003 he met King
Stephen I in Hungary and in 1008 he was in Kiev, under the protection of
Grand Duke Vladimir I who had been instrumental in introducing Christianity
to the Ukraine from Byzantium. Shortly before his death in 1009 — like
Adalbert, he was martyred - Brun wrote Adalbert’s biography. This was no
mere revision of the first life but a completely new version widi new facts and
stories inspired by the missionary fervour they both shared. Brun also com-
posed die \ita quinque fratrum of his spiritual master Benedict of Benevento,
who, along with his companion John and three young Poles, had been killed

44 Gerard of Csanad, Deliberatio , ed. Silagi.

45 Vita prior S Adalberti Pragensis episcopi , ed. Karwasinska; cf. Starnawski (1991), but also Kurbis (1991),
p. 243.

Cambridge Histories Online © Cambridge University Press, 2008

194

CLAUDIO LEONARDI

while preaching in Poland. 46 This text is an even clearer example of a mission-
ary hagiography of saints very recendy dead. In this form of hagiography auto-
biography plays an important, and often clearly stated, part. Of the sam & genre
but very different is the I Ita sancti Oudalrici by Gerard of Augsburg. 47 This
recalls die invasions of the Magyars, but the narrative is enriched by a number
of different elements including miracles and visions.

TROPES AND HAGIOGRAPHY

The intellectual centre of Europe still lay in France, Burgundy and Lotharingia,
where Carolingian culture had developed most fully. But other areas were
rapidly growing in importance, including the eastern areas of the old
Carolingian territories as far as Saxony, while Bavaria and Italy, especially
northern Italy, were becoming increasingly significant.

Many towns continued to maintain schools and be centres of learning; some
Carolingian centres of excellence disappeared, to be replaced by new centres.
Their rise and the weakened link between schools and culture meant that intel-
lectual life now depended less on schools than previously. Even though tenth-
century writings always show a local colouring, it was in this period that many
literary genres finally took on a clearly defined individuality; new genres were
also introduced. Writings began to be characteristic of their period rather than
of their region.

Sequences and tropes are a typical new genre. According to Notker of St
Gallen (c. 840-91 2), 48 texts and music not originally included in the canon of
the liturgy began to appear at the end of the ninth century; certainly the St
Gallen monk’s liturgico-literary invention was extremely successful. It con-
sisted of inserting a text into the vocalise which prolonged the singing of the
Alleluia in die Mass preceding the Gospel lesson: this is the sequence ( sequen -
tid), also known as the sequentia cum prosa or prosa. Following a tradition which he
claimed to have learned from a monk of Jumieges, 49 Notker regularised the
relationship between music and words (one note to a syllable) while maintain-
ing the autonomy of each. From this time on the genre underwent a great
development. 50 A trope is also a text produced for the vocalisation within a
liturgical text, but unlike the sequence it is not autonomous, since it also
includes some of the words of the liturgy; it can thus be inserted in die singing
at any point in the Mass. The earliest surviving tropers come from die tenth
century; by then the genre had already developed a full range of expression,

46 Passio sancti Adalbert ! , ed. Karwasinska; I Vita quinque fratrum, ed. Karwasinska; Dunin-W^sowicz

(1 972); Sansterre (1989). 47 Gerhard, Vita sancti Oudalrici episcopi Augustani, ed. Kallfelz.

48 von den Steinen (1948). 49 Ibid., pp. 1 54—5 .

50 von den Steinen (1946, 1947; 1967), pp. 1 1 5—50; Huglo (1987); Haug (1987).

Cambridge Histories Online © Cambridge University Press, 2008

Intellectual life 195

reaching its climax in the eleventh century. There are various kinds of trope:
the melogenous tropes of the Alleluia , which follow the liturgical text, logo-
genic tropes, which precede it, and meloformic tropes, which are interpolated
into a pre-existing text. 51

While this last kind died out before the end of the tenth century, the other
two were further developed and, together with the sequence, undoubtedly
greatly changed the liturgy experienced by the faithful in most of the monas-
teries and cathedrals of Europe. New texts and melodies introduced a previ-
ously unknown sense of solemnity to rites in the west. From St Gallen and,
according to well-founded hypotheses, from the region between Rhine and
Meuse (Mainz and Priim), perhaps even from Lotharingia (Gorze), the genre
spread to England (Winchester) and other areas of France - Autun, Limoges,
Aquitaine - and even to Italy. In the tenth century one of the logogenous
tropes was the famous trope dialogue between the angel who witnessed the
resurrection of Christ and the disciples who ran to the sepulchre: “Whom seek
ye in tire tomb, o dwellers in Christ? - The Cross of Jesus Christ, o dwellers in
heaven.’ 52 This eventually developed into the liturgical drama at the heart of
the medieval theatrical revival. Sequences and tropes soon achieved literary
and poetic maturity, partly because of the demand for them and partly because
of the stimulus of public performance on their authors. This maturity can be
appreciated in the dramatic trope written for one of the Masses of Christmas,
the Hodie cantandus attributed to Tuotilo of St Gallen (d. 913), the only writer of
tropes of the period whose name has come down to us. 53

Less connected with the liturgy were hymns, whose tradition goes back to St
Ambrose. 54 There are several large ninth-century hymn collections, probably
from St Gallen, Verona and Limoges. The habit of including older hymns in
later hymnals continued through the tenth century as tire Severinian hymnal,
better known as the Umbro-Roman hymnal, shows. 55 Throughout this period
the rhythmic line continued to alternate with the quantitative line, which it imi-
tated freely until the year 1000, by which time greater control of the structure
of die hymns had become more normal. 56

While die poetry of sequences, tropes and hymns is often of great depth
and high quality, the tenth century offers die first verse masterpieces of medie-
val Latin literature. Between the middle of the tenth and the beginning of the
eleventh centuries there was an outpouring of very significant prose and

51 See Corpus Troporum (1975—90), and also Jacobsson (1986), Silagi (1985), Leonardi and Menesto
(1990), Arlt and Bjorkvall (1993).

52 ‘Quern quaeritis in sepulchro, o Christicolae? — Iesum Christum crucifixum, o coelicolae’ See

Drumbl (1981); Davril (1986); Iversen (1987). 53 Jacobsson (1991). 54 Norberg (1954, 1988).

55 Hymnarius Severinianus, ed. Dreves; cf. Norberg (1977), but also Leonardi (1981).

56 Stella (1995).

Cambridge Histories Online © Cambridge University Press, 2008

196

CLAUDIO LEONARDI

poetry ranging in form from the instructive to the epic. One of the most
common forms of writing was hagiographic poetry, often confused with
historiographic poetry. When such a confusion occurs there is usually a ‘geo-
graphic’ transfer and die poem ceases to be classed according to where it was
produced and is subsequendy classed by its genre. As early as the beginning of
the Carolingian era Alcuin had written a life of St Willibrord pardy in verse
and pardy in prose, as well as die better-known X ersus de sanctis Euboricensis ecck-
siae , 57 Several typical Anglo-Saxon hagiographies of the tenth centuries
written in verse and prose have already been mentioned, and continental
European productions were not dissimilar. One of the main differences was
that in many hagiographies, especially in diose in prose, the hagiographical
details accompanied and were often mixed in widi historical details, because
the stories frequendy concerned people who had only recendy died, about
whom many facts were known that could not be omitted. These are hagiogra-
phies in which the transcendental component has become conventional, of
less interest than die story of the protagonists and their monastery and/or
bishopric. Contemporary hagiography inevitably tends to become historiogra-
phy.

The famous Life of Count Gerard of Aurillac, who died in 909, written by
Odo of Cluny around 925, can be considered the first hagiography of a
layman, although the saindy model applied to Gerard is still basically monastic
and hardly lay. 58 Hagiographies poured out of Cluny, from John of Salerno’s
hagiography of Odo in die middle of the century to Sirus of Abbot Maiolus at
the end of the century. 59 In these monastic hagiographies it is easy to see an
ideological element linked to the monastic reform movement associated with
Cluny, based on moral rigour, on distancing oneself from the political world
and creating a personal source of power, on prayer and on die celebration of
the liturgy. 60 Other centres of monastic reform also produced hagiographies,
the most significant being the X'ita Iohannis abbatis Gorgensis attributed to John
of Saint-Arnulf. 61

There was also a significant reaction against the power of Cluny and its hag-
iographic writings, for example in Adalbero of Laon’s rhythmus satiricus, an
invective against Count Landri of Nevers. Adalbero, born towards die middle
of the century, depicts the count, a member of Hugh Capet’s court, as an insid-
ious traitor. The twenty-eight Ambrosian stanzas of the poem dirow a shadow
over monastic life in Cluny at the time of Abbot Odilo and describe die monks

57 I, Deng-Su (1983); Alcuin, The Bishops , Kings and Saints of York, ed. Godman.

58 Odoof Cluny, Vita Sancti Geraldi A.urilacensis comitis libri quatuor, cf. Lotter (1983), Airlie (1992).

59 Iogna-Prat (1988). 60 Constable (1991).

61 John of St Arnulf, Vita Iohannis abbatis Gor^iensis; cf. Barone (1991).

Intellectual life 197

as warriors slaying their foes with the support of the pope. Anti-Roman satire
had already begun. 62

We have already referred to a number of Anglo-Saxon hagiographies and to
what might be termed ‘missionary’ writings from the eastern frontier of
Christianity. Other works of equal importance, though perhaps of less impor-
tance for the development of the ideal of sanctity, were also being produced at
this time: the Lives of two Saxon noblewomen: Queen Matilda, the wife of
Henry I, who died in 968 and merited two biographies 63 and the Empress
Adelaide, the widow of Otto I, who died in 999 and about whom Odilo of
Cluny immediately composed his Epitaphium. M

The most important characteristic of tenth- and eleventh-century hagiogra-
phy is its obsession with ‘territorial expansion’. Every Christian community
seemed to want a written life of the saint that best represented it and with
whom it could identify. 65 It was a phenomenon similar to, but different from,
the cities’ search for patron saints in previous centuries. People sought,
perhaps unconsciously, not only for a protector but for a model of saintliness
of some kind or other: monastic, episcopal, contemporary or ancient martyrs,
a queen or a simple monk. Given the obvious crises of the political institutions
of the time, which only towards the end of the century found any degree of
order and peaceful government, and then only in a few areas of Europe, and
given the crises facing the papacy and die clergy, the problems created by the
invasions from beyond the borders, and above all the effects of social and polit-
ical particularism, the blossoming of hagiography demonstrates die require-
ments of a number of attitudes which cannot simply be reduced to a need for
political and social security. There was also a search for a different code of
moral behaviour and greater spiritual awareness, a search which now looked
for historical, even ‘territorial’ points of identification. The hagiographer
therefore referred to saints and their real or supposed involvement with the
history of a particular town or city to produce a story which was understand-
able within die context of that particular place or city.

This deep connection between contemporary life, place and model of saint-
liness can be seen in the antiphrastic utterances of Letald of Micy. Speaking of
the miracles of St Maximinus, who lived in Micy during the sixth century,
Letald declares, ‘I am about to relate not tilings I have heard but tilings I have
seen’, 66 and of the vita of the early Christian Julian, bishop of Le Mans, Letald

62 Adalbero of Laon, Carmen ad Robertum regem , ed. Carozzi (for the other works see also the edition by

Hiickel); Brunholzl (1992) pp. 268—74; cf. Oexle (1978).

63 Vita Mathildis reginae antiquior, ed. Schiitte; 1 Zita Mathildis reginae posterior , ed. Schiitte.

64 Odilo of Cluny, Epitaphium A.dalheidae imperatricis , ed. Paulhart; Corbet (1986).

65 Hofmann (1991); I, Deng-Su (1991).

66 ‘neque. . .audita sed quae vidi narraturus sum’: Letald of Micy, Liber miraculorum Sancti Maximini

Miciacensis, col. 81 3D.

198

CLAUDIO LEONARDI

writes: ‘nothing pleases except what is true’. 6. In the tenth century a fantastic
hagiography is consciously trying to become reality.

This is one of the reasons for the appearance of die specialist hagiographer.
Some of these were authors of significance: Hucbald of Saint- Amand, whose
most important work is die l 'ita sancta Rictrudis; 6S Adso of Montier-en-Der,
who wrote a life of Clothilde among others ; 69 Theoderic of Fleury (or
Amorbach), who also produced a number of lives of Italian saints ; 70 Folcuin
and Fleriger of Lobbes . 71 Other audiors, mainly anonymous, were active in
many other centres, especially in the Frankish kingdom, Burgundy, and
Germany, for example in Trier, where Sigehard added a book of miracles to the
Life of Maximinus by Lupus of Ferrieres . 72 Ruotpert of Metdach related the
life and miracles of Adalbert of Egmont, but diere were also many others . 73
Great scholars like Radier of Verona also produced hagiographies: he rewrote
the Vita sancti U smart™ while Ekkehard of St Gallen wrote the Vita sanctae
Wiboradae™

One of the most important of these hagiographies, and a work of great lit-
erary merit, is the Miracula sanctae Fidis by Bernard of Angers, a student of
Fulbert of Chartres . 76 After a pilgrimage to the tomb of the saint in Conques,
Bernard decided to write about the miracles he had heard: cures, exorcisms,
conversions, but also the resurrection of animals, the freeing of miscellaneous
types of prisoner and the punishment and even deadi of pilgrims who failed to
make promised votive offerings to the saint. The tales are told in a free-tiowing
prose style dominated not so much by linguistic affectation as by a real enjoy-
ment of die narrative process and a dedication to the role of pilgrim extraordi-
nary.

Verse hagiographies are a minor product of the genre, often more stylisti-
cally polished but more abstract dian the prose versions. Apart from hymns
and odier short works, there are numerous true hagiographic poems diat
range from die Vita sancti Romani of Gerard of Saint-Medard 77 to the Vita
sancti Richard of Angilram of Saint-Riquier 78 and the Vita et passio sancti
Christophori of Waldier of Speyer . 79 The first and longest book of diis work
demonstrates the flexibility of this literary genre: Walther transforms his hagi-

67 ‘nihil placet nisi quod verum est’: Letald of Micy, 1 '/ita Sancti Iuliani , col 782B; cf. Cremascoli (1991).

68 I, Deng-Su (1990). 69 Werner (1991). 70 Poncelet (1908).

71 Dierkens (1983).

72 Lupus, 1 Vitae Maximini episcopi Trevirensis , ed. B. Krusch, and Sigehard, Miracula S. Maximini; cf. Zender

(1982). 73 Vita Adalberti diaconi Egmundae.

74 Rather of Verona, De Vita sancti Usmari\ cf. Golinelli (1991); but see also Dolbeau (1987).

75 Vitae sanctae Wiboradae, ed. Berschin. 76 Liber miracolorum sanctae Fidis, ed. Robertini.

77 Gerard of Saint-Medard, Vita Sancti Romani, PL 138, cols 171—84.

78 Angilram of St Riquier, Vita sancti Richard abbatis CentulensL, cf. Manitius (1923), pp. 533—5.

79 Walther of Speyer, Libellus scolasticus, ed. Strecker and ed. Vossen.

Intellectual life 199

ography into an autobiography, speaking of his own training and life and not
of those of his hero. This again demonstrates the literary need for hagiogra-
phy in the tenth century; the hagiographic genre served to identify the writer
with his subject.

THEATRE AND EPIC POETRY

Hrotsvitha of Gandersheim (c. 935-75), a nun, does not fit into die geographic
classifications of hagiography. Connected with the Saxon court, she was
perhaps the most famous and important of the women writing in the high
middle ages, and was above all the first real ‘love poet’ of medieval Latin
culture. 80 Her work can be divided into three groups: eight short hagiographic
poems, two historical epics and six celebrated plays. 81 Her hagiographies (all
written in Leonine hexameters apart from Gongolf in distichs) are not about
local saints but include apocryphal poems (on Mary and the ascension of
Christ) unconnected with any particular town or city, She also wrote die life of
the Merovingian saint Gongolf and the Cordovan martyr Pelagius (d. 925), the
lives of Theophilus and Basil of Cesarea, of Dionysius the Aeropagite and
Agnes of Rome. The texts are remarkable in that good always triumphs over
evil; even pacts with the Devil (as in the lives of Theophilus and Basil) cannot
prevent the saint from achieving communion with God.

Hrotsvitha’s hagiography is optimistic. Sin is overcome, the Devil does not
conquer, and pacts with him (the Faustian tradition in die west begins with
Hrotsvidia) cannot damn the soul of the sinner; she remains immune to all
these forces. God Himself would be a radier alien force if her own feminine
condition had not revealed His merciful nature to her. The discovery was
brought about through her writings on the Virgin Mary (the subject of the first
poem taken from the Apocrypha) and Agnes. In both poems the audior iden-
tifies with her heroine: Mary has made Hrotsvitha’s own choice and Agnes has
turned her back on romantic love in favour of a different, but no less intense,
form of love, that of the Heavenly Spouse, and in fact manages to persuade
her partner to share in this new love. Hrotsvidia’s saints are often lovers
(Agnes, and Proterio’s daughter in Basilius ) . Hrotsvitha’s hagiography is marked
by an enthusiasm for story-telling and a sympadietic attitude to die human
condition. She chose her own texts, put them into verse or wrote them herself
(as apparendy is the case with Pelagius) and devised her own narrative scheme
based on the conflict between good and evil and the final triumph of good and
God’s mercy. 82

80 Vinay (1978b), p. 554-

81 Hrotsvitha, Opera-, Vinay (1978b), pp. 483— 5 54; Dronke (1984b). 82 Schiitze-Pflugk (1972).

200

CLAUDIO LEONARDI

In her two epic poems, clearly written to commission, the final triumph of
good is not so obvious. The Gesta Ottonis celebrates the life of Otto I, while the
Primordia coenobii Gandeshemensis tells the history of her convent. The influence
of her sources and of the events which had to be included is large. It is interest-
ing, however, that the unifying force of the Gesta is not Otto himself but the
government of die world through various people (including Otto, Adelaide,
Berengar, Henry I and Liudolf) while in the Primordia the unifying force is the
monastery. But these forces merely impose an external unity on the historical
material and leave Hrotsvitha free to arrange die two poems into a series of
tableaux and stories in which she can give her optimism free rein, allowing her
once more to merge epic with hagiography . 83

Critics have identified her literary models as Virgil primarily, but also
Prudentius (fourth century) and Sedulius (fifth century), both Christian poets
of late antiquity. Her sources have also been identified quite easily as the Bible,
hagiography and the liturgy. It has to be said, however, that Hrotsvitha takes a
rather free approach to her sources, not in the sense that she ignores them but
that she works through them to give her work her own stamp. This is also true
of the work of Terence that she explicitly named as the source for her plays:
‘for there are others, who stick to sacred writings, and yet although scorning
other pagan writers frequently delight in Terence’s fictions ’. 84

The plays show even more clearly that her problem (both in life and art) was
how to equate human love with the ideal of Christian perfection, how to
describe the conflicts this produces and the solutions required. Her drama is
not contemporary as it is in die historical poems. Gallicanusis set in two periods:
that of the Emperors Constantine and Julian (fourth century); Dulcitius'vs, set in
the Diocletian period and Calimachus at the very beginning of die Christian era,
in the Ephesus of John the Aposde. 'Wrih Abraham and Paphnutius we move to
the initial, glorious monastic era of the fourth century, while Sapientia vs, situated
in Hadrian’s Rome. Hrotsvitha not only sited her ‘anti-Terentian’ dramas in two
different time frames, primitive Christianity and the fourth century, but also
used two different historical backgrounds, the persecution and martyrdom of
the Christians and the monastic life.

In the plays, even more than in her other works, history does not interest
Hrotsvitha; she is entirely wrapped up in her existence as a nun. History has
little significance in a nunnery and so is played down in her plays. She is inter-
ested in psycho-historical situations that could not have appeared in any drama
of her time. In an attempt to resolve the conflict between passion and perfec-
tion, Hrotsvitha tried to reduce its immediacy by situating her plays in the

83 Kirsch (1991).

84 ‘Sunt autem alii, sacris inhaerentes paginis, qui licet alia gentilium spernant, Terentii tamen fingmenta

frequentius lectitant’: Praefatii , ed. Homeyer, p. 233, ed. Winterfeld, p. 106 lines 3—6.

Intellectual life

201

distant past. In Calimachus and Abraham she depicts women in search of love
whose deep Christian faith and convictions mean that only God, the Heavenly
Spouse, can offer the type of love they seek. As a woman with a true monastic
vocation she exalts virginity and chastity and realises that death is often the
price that must be paid for the love of this terrible God who demands such
self-sacrifice from those who adore Him. Her plays usually end in martyrdom
and often deal with terrible carnal temptation (Drusianus in Calimachus) or
prostitution (Mary in Abraham and Thais in Paphnutius). This juxtaposition, at
times mechanical and superficial, offers, however, a solution to the problem
which was already visible in her earlier short hagiographic poems. A merciful
God allows men, or rather Christians, who are prepared to accept self-destruc-
tion to find a peace and serenity unknown to man: man is kind to woman and
the old hermit Abraham is finally able to express all his affection to his niece
Mary.

Hrotsvitha’s greatest literary achievement coincides with her finest intellec-
tual intuition: monasticism no longer consists in ivory tower contemplation
and aristocratic and imperial connections, but is open to mankind and focuses
greater attention and love on man. Hrotsvitha perceives man as love, even
though she is unable to show die love between man and woman openly. In
some plays there is only a female protagonist because woman is the one true
ally of God in the conversion of men, but in Abraham both man and woman
widi their delicate tenderness and affectionate outpourings are protagonists
and the keyword of the drama is pietas, ‘divine piety, which is greater than all
created things’. 85 In this sense Abraham is Hrotsvitha’s masterpiece. 86

The greatest of the epic poems written in die tenth century is Waltharius. But
it was not the only epic poem produced. In the first half of die century an
anonymous author in northern Italy wrote a panegyric on Berengar, which is
principally a description of the struggles between Berengar and Wido of
Spoleto. The Gesta Berengarii, composed in 1,090 rather elegant hexameters, has
numerous hellenisms and many references to Virgil, Statius, Prudentius and
Sedulius. Unusually, the audior supplied his own glosses. 87 While the anony-
mous Gesta in its celebration of contemporary power is close in spirit to
Hrotsvitha’s Gesta Ottonis, her Primordia was perhaps the model for other
poems celebrating monastic and episcopal life such as the Gesta Witigowonis of
Burchard of the Reichenau. 88

Two epic poems of a different type each represented die beginnings of a real
literary tradition. The Gesta Apollonii, possibly written in Tegernsee, tells the life
and adventures of Apollonius of Tyre and is a reworking in hexameters of a

85 ‘superna pietas maior est omni creatura’: A.braham vii, i o, ed. Homeyer, p. 316, ed. Winterfeld, p. 158,

line 4. 86 Vinay (1 978b), pp. 5 32—5 3 . 87 Gesta Berengarii imperatoris , ed. Winterfeld.

88 Purchard, Gesta Witigowonis , ed. Strecker; cf. Autenrieth (1985), pp. 101— 6.

202

CLAUDIO LEONARDI

Latin prose story of the fifth century, in its turn a Christianisation of an old
Greek text. 89 Even more successful were epics about the animal world, such as
the Ecbasis cuiusdam captivi per tropologiam, the first epic of its kind in die middle
ages. In classical times satirical writers from Aesop to Phaedrus had made
animals the main characters in their fables. This tradition was taken up again -
for example by Sedulius Scottus - in die Carolingian period and reached a high
point in the eleventh century with the work of Adhemar of Chabannes
(988-1034). 90 But until the Ecbasis no-one had produced a whole story about
animals in a single narrative that was neither instructive nor satirical but epic; it
is this epic structure, even though fragile and subtle, diat holds the poem
together. Consisting of 1229 Leonine hexameters, composed perhaps in
Lotharingia towards the end of the tendi century, it is really two stories, one
within die other. Although the meaning of the poem is enigmatic, its literary
worth is undeniable. The anonymous writer of the only successful comedy
written in the early middle ages gives us a series of well-described scenes por-
traying animals and characterising them figuratively and psychologically (with
references to Horace). 91

Waltharius (a poem of 1,453 hexameters) is poetry of a rather higher
calibre. 92 Its dating is still disputed, but opinion now tends towards the middle
of the tenth century radier than die earlier dating to the first half of the ninth
century. The revised dating also means that Ekkehard I of St Gallen (910-73),
known to have been audior of a Vita Waltham manufortis , has now been put
forward as its author. 93 Although Dieter Schaller has recently claimed that
Ekkehard’s authorship has not been proved beyond doubt, 94 the likelihood of
the poem’s having been written in St Gallen and therefore being attributable to
Ekkehard has been strengthened by Schaller’s own conclusions: there is no evi-
dence for the existence of a Waltharius poem in die Carolingian period, while
there are strong links between early eleventh-century Mainz and Ekkehard IV
of St Gallen, who undoubtedly knew and reworked a Vita Waltharii written by
Ekkehard I. 95

The dieme of Waltharius is neither Christian nor classical - this is a story in
the Germanic tradition. By the high middle ages the cultural heritage of the
Germanic peoples had already become the subject of Latin literature, with
masterpieces such as Bede’s Elistoria ecclesiastica and Paul the Deacon’s Ehistoria
Langobardorum telling die great tales. These were works of history and not epic

89 Gesta Apollonii , ed. Diimmler and ed. Ermini.

90 Adhemar of Chabannes, Fabulae , ed. Gatti and Bertini.

91 Ecbasis cuiusdam captivi per tropologiam, ed. Strecker; cf. Gompf (i 973).

92 Waltharius , ed. Strecker; cf. Langosch (1973); Onnerfors (1988).

93 Brunholzl (1992), p. 63. 94 Schaller (1991), p. 437.

95 Ibid., p. 436. Cf. also Werner (1990), pp. 101-23.

Intellectual life

203

poems. In the Carolingian period the Germanic tradition had found a voice in
Karolus Magnus et Leo papa-, but Waltharius was something new and different: it
was no longer a single poem but an entire story, that of Walther’s flight and
return to his homelands.

There was probably an oral Germanic version of the story in existence
before die appearance of the Latin text but the theory that the Latin is a later
version of a German text can now be excluded. What is clear, however, is that
the Germanic tradition itself is die subject of the poem although the refer-
ences to Virgil in the flight theme, in the description of the battles and in the
sense of adventure which runs through the work are obviously important, as
are the references to Prudentius. Hrotsvitha’s optimism is also present in the
poem, with a vein of Christianity tempering the Germanic tale. The main influ-
ence is Carolingian poetry with its occasionally rather laboured metric struc-
ture and somewhat creaky imagination. But from this structure Ekkehard (or
the anonymous author from St Gallen) has created a broad poetic narrative
that allows the Germanic epic to make a welcome entry into die Latin-
Christian tradition. Walther is not on a great mission: this is the story of man
who has reached maturity and full possession of all his powers returning home
after many adventures to die woman he loves (Hildegund is a shadowy figure
whose only real function is to represent a haven of security for the hero). It is
the story of a man who is strong and heroic, but who shares the anxieties, die
uncertainties of travel, the fear of the night, of the unknown and of the enemy
with his friends/enemies, that is with other men. Walther seems almost over-
come by the trials of life and die final battle and its uncertain outcome (because
Walther knows that he may be the loser) are at the heart of the poem. Walther’s
best quality is not his strength but his desire for peace and light which - like
Hrotsvitha’s merciful pietas— runs through the work:

Behold, however things may turn out, here I shall lie
Until the revolving sphere brings back the longed-for light,

Lest the land should say ‘That king, the proud king,

Has stolen a thief’s flight through the shadows, as is his wont .’ 96

Waltharius marks the beginning of the great Germanic epic tradition. 97

The tenth century produced another surprise. Far from St Gallen, in Micy in
the Loire Valley, a monk named Letald (c. 945-96) wrote a short farcical epic

96 En quocumque modo res pergant, hie recubabo,

Donee circuiens lumen spera reddat amatum,

Ne patriae fines dicat rex ille rex superbus
Evasisse fuga furis de more per umbras.

Waltharius , ed. Strecker, p. 71, lines 1 1 51— 4

97 Von den Steinen (1952); Vinay (1978c); Dronke (1984a).

204

CLAUDIO LEONARDI

poem: a tragedy with a happy ending, but in this case the ending is a comic
parody. In Within piscator, X ersus de quodam piscatore, quern ballena absorbit, Letald
tells how the Englishman Within (he who is within) one day goes fishing and
ends up inside the belly of a whale. He tries desperately to escape but in vain,
until he remembers that he has a knife and with this he keeps slashing at the
whale’s stomach, wounding it ever more gravely until finally, after several days
imprisonment he reaches the heart and kills the fish, which beaches on the
coast he set off from. The reference to Jonah in die Bible is obvious.
The ending, however, bears no relation to the biblical story. The inhabitants
of die village run down to the beach to see the whale and open it up and share
out die meat. Within cries for help and is taken for the Devil. The villagers then
organise a propitiatory procession and an exorcist, but in die end they recog-
nise the fisherman and all ends happily . 98 In Within Letald gives us a pleasant,
fantastical story that is almost a joke and which parodies both the high-minded
ideological example of the biblical Jonah and the committed epic poetry about
kings, queens and knights.

HISTORIOGRAPHY AND LITERATURE

Intellectual production during die whole century was notably historiographic.
With its hagiography and its epic-historical and biographical poetry, the period
is best characterised by its interest in contemporary and past events and its
varied historiographic output. It is as if the enormous quantity of historio-
graphical writing by often anonymous authors answered a need. This was no
longer the need to rediscover a cultural past as the Carolingian era had redis-
covered Latin and the works of the auctores and Christian writers, rather the
need to rediscover the history of a people and to regain a lost sense of cultural
awareness and responsibility. No other literary genre of the tenth century pro-
duced so many works as historical writing, from annales to chronica , from gesta to
biographies and autobiographies, a form which made its first medieval appear-
ance in this period.

The annales formula, ennobled during the Carolingian period by its pre-emi-
nence in court literature, once more became the accepted form of memoirs,
especially in the monasteries. Talented writers such as Flodoard of Rlieims had
no hesitation in calling dieir historiographies annales. At least thirty annalistic
works date from the tendi and beginning of the eleventh centuries: many come
from eastern Europe, including Bavaria and Saxony, but several odiers were
written in northern France, in the valleys of the Loire and the Rhine, and a few
were produced in other places like Spain and Italy . 99

98 Letald of Micy, Within piscator ; ed. Bertini; cf. Berdni (1991). 99 Hofmann (1991).

Intellectual life

205

Flodoard’s Annales follow die tradition of chronicling events year by year . 100
But he also innovates, and his chronicles are rich in details which show a design
tending more towards the narrative than towards the documentary and an
alternation of local news with news of more general interest. In Flodoard’s
writings, the annalistic genre, which continued to enjoy great popularity for a
number of centuries, acquired several structural features which extended its
meaning but also brought it to the brink of extinction . 101

Apart from his work in continuing a great cultural tradition, Flodoard is also
remembered for two other, equally innovative works: the De triumphis Christi
and the Historia Remensis ecclesiae. The first is an apparently incomplete hagio-
graphic epic poem, which recounts the exploits of the saints of Palestine,
Antioch and Italy. The work sets out to be a legendary of die universal church
and is innovative in its hagiographic scope because of its historico-geographic
structure, which extends out from Palestine to the whole Christian world.
Flodoard can be considered the historian and hagiographer par excellence of the
period. Like many other writers of the time, he considers historiography and
hagiography as the two components of historical conscience: history as a series
of events and personages and history as sublimated in the viri Dev. human
history and divine history that together lead to an understanding of the times . 102
Flodoard’s Historia also starts another tradition, or at least reinforces an existing
one. He uses public and archival documents such as epigraphs, private charters
and letters as a further basis for his historical narratives, following the model of
Agnellus of Ravenna’s Liber pontificalis and John Hymmonides’ Xlta Gregorii
from the first and second half of the ninth century respectively . 103

The chronica genre also continued to be employed both in bishoprics and in
monasteries, but its heyday came later in the eleventh century. Nevertheless,
the tenth century also produced several great chronica-. Widukind of Corvey’s
Res gestae Saxonicae, the story of the house of Saxony; Benedict of St Andrew by
Monte Soracte’s Chronicon, about Rome; and an anonymous south Italian
writer’s Chronicon Salernitanum. The last of these is a local chronicle of facts and
extraordinary happenings . 104 Benedict describes Rome and its surroundings
from a monastic point of view, which imposes a distance on events, as if seen
from another form of reality . 105 Widukind, by contrast, is aware that he is
telling a story about power; he has no illusions about the world he is describing
and no time for the unreal . 106 All three chronicles share one feature, however:

100 Flodoard, Annales, ed. Lauer. 101 Sot (1993).

102 Flodoard, De triumphis Christi ; Jacobsen (1 978) .

103 Flodoard, Historia Remensis ecclesiae , ed. Stratmann.

104 Chronicon Salernitanum , ed. Westerbergh; Oldoni (1 969).

105 Benedict of Soracte, Chronicon , ed. Zuchetti; Oldoni (1991).

106 Widukind, Res gestae Saxonicae, ed. Hirsch and Lohmann; Beumann (1950).

2o6

CLAUDIO LEONARDI

the historian identifies entirely with the place he is writing about (monastery,
city or kingdom). In consequence universal chronicles, such as Regino of
Priim’s Chronica sen libellus de temporibus dominicae incarnationis completed in
908, 107 are of litde interest to him.

Two historians stand out: Richer of Rheims and Liudprand of Cremona.
Richer, who had studied with Gerbert, was still alive at the end of the century.
The heir to a great historiographic tradition, his conception of the art was
completely different from Flodoard’s. Richer’s writings follow no chronologi-
cal or any other apparent order, and seem determined only by his own opinion
of the history of his time. In selecting die contemporary events to record in his
Historiae, Richer makes no use of existing sources or judgements and shows his
skill in using literature as a basis for his historiographical texts. 108

Richer was following in the footsteps of Liudprand (c. 920-72), one of the
greatest stricdy literary geniuses of the century, whose achievements dwarfed
Richer’s. Liudprand had studied and was perhaps born in Pavia, but he did not
dedicate himself to a life of study. Instead, he made an early entry to Berengar
II’s court and later, controversially, left for the Saxon court of Otto I, who
created him bishop of Cremona. He continued in the emperor’s service,
however, and in 967 went to Constantinople to request the hand of
Theophanu for Otto II (having already been to Constantinople previously on a
mission for Berengar). 109

Liudprand’s historiographic works all have contemporary settings and often
drift into autobiography: the Antapodosis or Liber retributionis regum atque princi-
ple m Europae ; the Liber de rebus gestis Ottonis magni imperatoris, and the Relatio de leg-
ations ConstantinopolitanaA 0 In the last two books die titles point to his role as
Otto’s counsellor and ambassador but he is also present in the first (and most
famous) of die works. Liudprand’s historiography abandons all pretence of
objectivity, even the apparent objectivity of the various types of annal, chroni-
cle and history which try to present a list of facts in an attempt to conceal the
historiographic and ideological selections made. Liudprand conceals nothing
and describes his facts with a crudity and immediacy unknown in Carolingian
historiography. His stories do not only or even principally describe greatness
and nobility of intent and action; they are more concerned with their brutality,
unmasking foul intentions, dirty affairs and the vulgarity and obscenity of rela-
tions, including sexual relations, between men. If a Christian God exists in his
writings, His actions are mechanical, so that Liudprand, of Lombard descent,
seems to be reintroducing a primitive Germanic spirit into the literature of the
period.

107 Regino, Chronicorr, Schleidgen ( 1977;. 108 Richer, / fistoriae, Kortiim (1985).

109 Lintzel (1933); Sutherland (1988). 110 Liudprand, Opera', Chiesa (1994).

Intellectual life

207

The unrivalled greatness of his historiographic writings stems from his
despair at finding himself impotent in the face of events, and from his obsti-
nate, furious reaction to events he has understood or believes he has under-
stood. From the literary point of view, this gives an aggressive edge to his
narrative, which increases the more as he manages to convey anger in his narra-
tive-fictional interpretations of events which he has heard at third-hand and
which may be either truth or fiction. Liudprand produced a number of unfor-
gettable realistic portraits and tableaux. His view of history was anti-epic and
he regularly de-epicised every event: epic in reverse. His portraits such as those
of Nikephoros Phokas or of Marozia and Theodora, the libidinous debauch-
ers of Rome, or of Berengar and Willa are justly famous. His portraits and tab-
leaux are generated by an autobiographical style of writing, which
contemplates the meaninglessness of history with its irrationality and earthly
brutality and uses descriptions to expose this. 111

Other notable authors besides Liudprand were working in Italy in the Po
valley in die tenth century. They included Atto of Vercelli, the author of a
commentary on the Epistles of St Paul and especially of a Polypticum quod appel-
laturperpendiculum, an extremely obscure political treatise which is also a politi-
cal satire and compendium of wisdom. 112 The Ottomans also imported
teachers and codices from Italy (not only north Italy), often via Milan. Gunzo
and Stephen of Novara were at Otto I’s court at the same time as Liudprand, 113
while the manuscripts from Otto Ill’s library, still in Bamberg, include works
by contemporary Italian authors such as Eugenius Vulgarius. 114 Mainly from
southern but also from northern Italy, die texts on Roman history were clearly
used by the Ottomans to appropriate imperial ideology. Both John
Philagathos, the bishop of Piacenza (later abbot of Nonantola) and John, the
German bishop of Vercelli, played important roles in this ideological trans-
fer. 115

The cultural relations between Italy and Germany, two countries relatively
little touched by Carolingian culture, cannot explain die quality of Liudprand’s
prose (which made use of prosimetrum in the Antapodosis and inserted poetry
into die Relatio) n 6 or the prose of the odier genius of die period, Rather of
Verona (c. 890-974). Rather came from Flanders and was educated at Lobbes.
Like Liudprand, he was an autodidact, although he had received an excellent
scholastic education. He too was a bishop, of Verona, and was twice expelled
from his see by clergy and politicians who disagreed with his policies, only to
be later reinstated. 117

Like Liudprand’s, Radier’s personality was bizarre and eccentric, and so

111 Vinay (1978a), pp. 391— 432. 112 Atto of Vercelli, Polipticnm, ed. Goetz; Wemple (1979).

113 Ferrari (1991), pp. no— 14. 114 Pittaluga (1991). 115 Ferrari (1991), pp. 1 10— 14.

116 Jacobsen (1985); Scherbantin (1951). 117 Jacobsen (1989).

208

CLAUDIO LEONARDI

was his prose: it is like Liudprand’s in being rich and full of rare words and
apparendy irregular constructions, but it is also very different and not strictly
comparable. Liudprand narrates a story which happens to include his first-
hand experience; Rather narrates himself, though he hides this behind moral or
legal debates about the nature of hostility towards him. Liudprand recounts a
story and its action; Rather is defeated by a problem he never explicitly admits,
die difficulty of reaching truth and of making it reflect reality. As a result
Radi er is a lonely character and die consciousness of unbearable solitude is the
diread running through all his works from Praeloquia to the Qualitatis coniectura
cuiusdam, to Phrenesis and the letters . 118

Recendy the influence of the Consolatio of Boethius has been pointed out in
the works of the older Rather and his junior Liudprand (especially the
Praeloquia and Antapodosis) as the Phrenesis is written in a prosimetric structure
like the Consolatio and the De nuptiis of Martianus Capella. It has been claimed
that Rather, and later Liudprand, took Boethius’ concept of seeking consola-
tion widiin oneself for the antagonism of the powerful and used it as the
model for radical satirical criticism of the moral and political condition of the
time . 119 This may define the literary genre to which they belonged but it can
only offer a limited explanation of their literature and of die spiritual meaning
they both gave to their lives and works.

In describing Rather’s Praeloquia as a book in which ‘whoever reads it will
find many things while reading which can provide as much pleasure as profit to
the minds of those reading ’, 120 Liudprand is speaking less of Rather than of
himself; his own view of history finds a way to salvation in writing and narra-
tion, the means of achieving inner calm and taking pleasure in life. Rather, by
contrast, cannot do this; in his work there is the consciousness of an absolute
solitude which can be described but not resolved through writing, for the reso-
lution of his predicament would have to come through an absolute truth he
knows he can never express. As a result, Rather’ despair gives rise to die first
true form of ‘autobiographical representation in the high middle ages ’ 121 but it
also leads to an absolute desire for existence beyond history, which history
cannot satisfy. He expresses this desire through his hope and fear that history
will end in the millennium, allowing him to escape from the cruelty and despair
of history into true metahistory.

The period between the ninth and eleventh centuries was rich in many other
forms of intellectual activity and saw the rise of many other literary forms,
both old and new. But above all, the period was characterised by its sense of

118 Vinay (1989a, 1989b). 119 Staubach (1991).

120 ‘Quem si qui legerit, nonnullas ibi hac sub occasione res expolitas inveniet, quae legentium intellec-

tibus non minus placere poterunt quam prodesse’: Liudprand, A.ntapodosis \\\, 52, p. 101.

121 Vinay (1989a), p. 135.

Intellectual life

209

historical awareness. Hagiography, biography and true historiography, in all
their various forms, sprang from this basic realisation of the need to match
oneself against history.

One of the greatest writers of the age, Gerbert of Aurillac (940/ 50-1003)
was also one of the most aware of the period’s historical context. While other
writers contented themselves with descriptions of the past and/or the present,
or invented the legends of men, such as saints, who were historical representa-
tions of perfection in God, and while Rather and Liudprand measured them-
selves against history and came away defeated after having merely touched it
through their writings, Gerbert dominated and took control of history, deter-
mining its course. Gerbert too had known defeat, but not like Rather in Verona
or Liudprand at the hands of Berengar. Gerbert was a victor and in him we
have an extraordinary phenomenon: the intellectual who holds die reins of
power, literature as power. While other great writers carried historiography to
its greatest literary peaks, Gerbert plunged it into the world of politics.

Gerbert was born in the Auvergne and became a monk in Aurillac. He
studied in Catalonia, where he came into contact with Arab learning, and Pope
John XIII called him to Rome where he met Otto I. He left Rome for Rheims
to become director of its school. In 9 8 1 he was back in Rome and in Pavia Otto
II created him abbot of Nonantola. He won a famous disputation with Otric
of Magdeburg before the emperor in Ravenna, but Otto II died in 983 and
Gerbert returned to Rheims. On the death of Archbishop Adalbero in 989 he
succeeded to die see, but his tide was not confirmed by the pope. He therefore
left Rheims for the court of the Saxon emperors and became tutor, but mainly
counsellor, to Otto III, who was still a minor. Otto had him enthroned as arch-
bishop of Ravenna and in 999 as pope. Gerbert, the first French pope,
assumed the name of Sylvester II and enjoyed a relationship with Otto III
similar to Sylvester I’s with Constantine: that of the greatest ecclesiastical and
temporal powers in the western world working together in close harmony. 122

Gerbert was no historiographer, but both a political intellectual and a great
writer. His work shows how far Carolingian culture had now been left behind.
He was a master of the arts of the trivium and his travels to Italy were partly
made in order to find the works and stimuli needed to complete his training in
logic, as he was not satisfied with the few Ciceronian references in Carolingian
rhetoric and dialectic. His extraordinary scientific, technical and practical work
reinjected arts of die quadrivium from Spain into the general culture of the time.
France had been the centre of Carolingian culture but Gerbert makes it clear
that innovation now comes from outside France, from the Arabs and from the
traditions still maintained in Italy.

122 Gerbert of Aurillac, Opera', Opera mathematica , ed. Bubnov; Episto/ae, ed. Weigle; Gerberto ( 1985 ).

210

CLAUDIO LEONARDI

In his letters Gerbert’s successful prose style is consciously modelled on that
of die ancient auctores and is careful and controlled but effordess, far from the
linguistic contortions of Radier and Liudprand, for all their literary effective-
ness. He is immersed in writing and culture, ‘in study and politics we teach what
we know and learn what we do not know’; 123 though in politics as well, as the
sentence just quoted makes clear. Not surprisingly, Gerbert was believed to
have made a pact with the Devil: his successful and glorious career could only
have been achieved with Lucifer’s aid. 124 Nor is it surprising that his vision of
God has much in common with Boethius, perhaps his most intiuential model.
For Gerbert, God is total intelligence and perfect understanding of the world
because the world can be controlled by the intellect. 125 In this respect Gerbert’s
views are diametrically opposed to those of Rather. Both are ill-at-ease with
the theory of the perfect unity of truth and history, but while Radier considers
it impossible, Gerbert believes the intellect could grasp the concept and use it.

THE MILLENNIUM

The most striking feature of the era is the eschatology provoked by the immi-
nent arrival of the millennium: history must be told and should be understood
and directed but it can also come to an end. The great intellectual and spiritual
inheritance of the time is the understanding that history may have no future
but, in so far as it exists, is a function of the future. Beatus of Liebana,
Ambrosius Autpert and Aimo of Auxerre between the eighth and the end of
the ninth centuries had all interpreted the Apocalypse of John the Aposde as a
message to the individual, free of any historical context. In the tenth century a
different interpretation of the book produced a fear of the millennium and the
possible end to history that might accompany it.

It was one of Gerbert’s friends, Adso of Montier-en-Der (910/1 5-992) who
reintroduced the eschatological reading of die Apocalypse. In his widely read
Epistola de ortu et tempore Antichristi dedicated to Gerberga, Otto I’s sister, Adso
describes the Antichrist as a person, the son of Satan, who at the end of the
millennium is freed from his chains and brings history to an end. Adso voiced a
common tension: the fear of, but also the longing for, an end to history. 126

Through Augustine, and in particular his De dvitate Dei , the west perceived
history as a single process produced by the war between God and the Devil,
and understood die one thousand of the Apocalypse to mean the historical
millennium. The end of die world was perceived as a historical event following

123 ‘in otio, in negotio, et docemus, quod scimus, et addiscimus, quod nescimus’: Gerbert, ep. 4, p. 73.

124 Oldoni (1977, 1980, 1983). 125 Riche (1991), p. 421.

126 Adso of Montier-en-Der, Epistula ad Geihergam reginam de ortu et tempore Antichristi , ed. Verhelst;
Konrad (1964).

Intellectual life

21 1

a last battle in Jerusalem between the Christ and die Antichrist. This possibly is
why Adso, already very advanced in age, took ship for the east in 992, dying at
sea after five days. The desire to see the end of history was also a desire to see
the final outcome of the last batde and the ultimate victory of Christ for all
men.

The great legacy of the time is its view of the future. This formed the basis
for thinking in the west for many centuries to come: not only the expectation
of the end but also the knowledge that Christian perfection is to be found
within history. In this sense the works of Gerbert and Adso are both parallel
and convergent.

CHAPTER 8

ARTISTS AND PATRONS

Henry Mayr-Harting

ARCHITECTURE

The architectural world of the tenth and eleventh centuries is not easy to
recover either from what is now to be seen or from the literary sources. First,
although we have the literary and archaeological evidence for Ottoman royal
palace complexes such as those in Magdeburg or Ravenna, our evidence is pre-
dominandy that of churches. And second, as to churches, we see some remark-
able experimentation in what has survived, but die literary texts, our only
evidence for so much that has been destroyed or rebuilt, tend to present build-
ing by abbots and bishops in its traditionalist aspects. Often we have to work
from analogies. The church of Romainmotier in modern Switzerland is prob-
ably our best chance of seeing what die second church of Cluny looked like,
and Nivelles, where around 1000 a cousin of Otto III called Adelaide was
abbess, and where the church has a transept at each end of the nave, is likely to
redect the appearance of Bishop Notker’s (972-1008) cadiedral at Liege.

One church of novel character which can still be seen is that of St Martin de
Canigou, dating from die earliest years of the eleventh century, and built under
die patronage of Count Wifred of Cerdana. The Pyrenees, as Puig y Cadafalch
long ago showed, was an important region for the early development of
Romanesque styles. It would be a great mistake to regard this region as out of
die way, despite the impression of remoteness die monastery of Canigou now
gives, standing on a magnificent spur of the mountain of that name, and com-
manding staggering views upwards from its cloister. For it was close to the
great route which linked Spain and the Mediterranean to the heart of rich and
productive Lodiaringia via the rivers Meuse, Saone and Rhone. Canigou is a
traditional, ‘first Romanesque’ church in its use of small, unsmoodied stones
and its triple-apse plan, but its remarkably slender pillars support a daring
barrel vault over the nave.

However, the chief object of this discussion is not to engage in an analysis
of stylistic development, but rather to ask how the ecclesiastical architecture of

212

Artists and patrons

213

the period is related to what went on inside the churches. Carolingian monasti-
cism, with its interest in Roman and eastern chants, was already more liturgised
than the Rule of St Benedict had envisaged, and tenth-century monasticism,
whether we consider the so-called Gorze Reform, Cluny and Fleury, or the
English Benedictine revival, heightened this tendency. We know one of the
raisons d’etre of the various tribunes, the west work and the separated areas of
Agilbert’s church of Charles die Great’s time at Saint-Ricquier, for it is clear
from his Ritual Order. It was to enable choirs of monks and boys, separated
from each other, to answer each other antiphonally across the church with
impressive echoes, rather as the various tribunes of the dome of St Mark’s,
Venice, were intended to echo with the trumpets of Gabrieli’s music. One
cannot understand Carolingian churches, for all the importance of saints’
shrines and relics in their lay-out which has been rightly stressed, without
recreating for oneself the sounds which they were intended to contain. It is
quite as much so with the churches of the succeeding age. The musical stave
was only invented in the eleventh century, and it is clear that choirs of monks,
nuns and canons were still in the ninth and tenth centuries expected to know
their musical notes by heart, and the neums which positively sprouted in the
chant books of this age are considered to represent die movements of a con-
ductor’s hand, as well as reminding the choir of melody, and indicating rhythm
and ornamentation. An inspired precentor must have been able to elicit dra-
matic effects from his schola cantorum. We know that Liege was famous for its
chants under Bishop Stephen (903-20), that it was an experience to hear the
chants on Christmas night in Trier Cathedral or in the open air at Augsburg on
Palm Sunday, that hymnals spread far and wide from the abbey of St Gallen
(the Solesmes of die day), and that a spate of Marian andiems began to be
composed at Reichenau at die latest in the early eleventh century which appar-
endy gave us among others the Salve Regina. The tenth-century English show a
near craze for building organs, whose main function must have been to accom-
pany the chants; the evidence is mosdy English, but one may doubt whether
the phenomenon itself was so confined.

One notable musical composer was Odo of Cluny. While he was abbot of
Cluny (926-44) he was approached by the monks of his former monastery of
Tours to write some longer antiphons for Saint-Martin than those they had,
whose length might relieve them ‘of the monotony of repeating diese very
short ones’. He praised the brevity of their antiphons and expressed disgust at
the prolixity they demanded; they warned him that he would displease Saint-
Martin if he refused and that his excuse signified a hidden pride. Odo gave way
and composed antiphons ‘in which the meaning and sound agreed so well, that
it seemed that nothing could be added or taken away from the sense, nothing
found more sweet in the modulations of the melody’. His biographer adds that

214

HENRY MAYR-HARTING

they were retained to that day in Benevento, not an uninteresting observation,
since it shows the power of music to unite a politically fragmented world, as
between Liege and St Gallen, Tours or Cluny and Benevento. 1 There is much
evidence for die interest in musical dieory at this time, derived above all from
the writings of Boediius, and Odo is part of it, but this story shows that he was
no mere musicologist.

We have unfortunately litde hard evidence of how the various parts of a
tenth-century church were exploited musically, though diat does not invalidate
hypotheses drawn from the nature of the architecture itself, but there is a sug-
gestive story from amongst the tenth-century miracle narratives of St Faidi at
Conques about the cure of a blind man called Gerbert. In the night of die vigil
of the feast of St Michael, St Faith appeared to this man in a dream and told
him to join the procession of monks to the altar of St Michael the following
day after Vespers where God would restore his sight. And so it was done.
Gerbert accompanied the monks’ procession to die oratory of St Michael,
where, while they sang the antiphon in honour of the coming festival, the hea-
venly artifex or creator enabled him to see. 2 From the account it is clear that we
are dealing with a kind of Westwork, whether of die tenth century or the ninth
we cannot tell, of which the oratory of St Michael constituted the third level,
above a solarium which was itself supported by a vaulted structure on the
ground door. It is not clear, therefore, if the oratory gave onto the church as a
second level must have done, but it is quite likely, since we also learn from the
account that above the oratory was yet a fourth level with bell-tower (all this of
course disappeared in the great rebuilding of the eleventh and twelfth centu-
ries) . The important point for our purposes, however, is the use of the high gal-
leries for music.

A church of striking originality for its galleries along the length of the nave
at triforium level is Gernrode in the heardands of tenth-century Saxon rule.
Close to the royal nunnery of Quedlinburg with its staggering views across to
the Harz Mountains, Gernrode was founded as a nunnery in 961 by Margrave
Gero, one of the great military commanders of Otto I on his eastern frontier
widi die Slavs. Having lost his son Siegfried in the Slav wars, Gero founded
here a house of canonesses with his widowed daughter-in-law, Hathui, as dieir
abbess. Hans Jantzen, in his brilliant analyses of the aesthetic of Ottoman
churches, wrote of die strong rhythmical counterpoint between the arches of
this triforium and die great openings of the nave arcade underneath it, and he
developed the pleasant conceit diat die austere and noncommital exterior of
the building, in contrast to its lively and solemn interior, was just like the taci-

1 John of Salerno, Vita S Odoms abbatis Cluniacensis , c. io, col. 48c.

2 Liber miraculorum sancta Fidis 11, 1, pp. 91—2; see Lehmann-Brockhaus (195 5), pp. 27—8.

Artists and patrons 1 1 5

turnity of its founder himself as described by Widukind. 3 Be that as it may, one
cannot look at diese remarkable galleries without supposing that die nuns,
whose understanding of their liturgy was so litde staid and conventional that
soon after die foundation of Gernrode they gave us our first known commen-
tary on the psalms in the Old Saxon vernacular, would not sometimes have
tested their musical possibilities. Anyone who has tested the acoustics of the
west work in Archbishop Brun of Cologne’s (953-65) great church of St
Pantaleon will know that the cool well of space contained within the solemn
rythms of its arcaded galleries must provide a veritable echo chamber for a
choir of monks or boys. Here is another church whose building was initiated
by a patron close to Otto I, in this case his brother. The building at St Pantaleon
was continued in a later campaign apparently stimulated by the Empress
Theophanu, wife of Otto II, around 984, from which period some exceptional
fine stone sculptures have been discovered, including a head of Christ and
angel reliefs, which probably adorned the west facade. 4 The nave of Bishop
Bernward of Hildesheim’s (993—1022) great early eleventh-century church for
the monastery of St Michael, Hildesheim, has no galleries in the nave, but the
galleries at two levels in the transepts, together with die majestic vistas of the
interior and arrangement of external towers, must have contributed to the des-
ignation of this church by contemporaries as a templum angelicum.

This was a very angel conscious society, as one sees from the often domi-
neering postures of angels when they appear in art and literature. Whether the
monks and nuns of whom we have been writing always wanted to live like
angels may be doubted, but when one looks at their architecture there is a
stronger case for thinking that they sometimes wanted to sound like angels.

The liturgical efflorescence of the tenth century partly found expression in
small dramas, which were now for die first time fitted into the structure of the
monastic liturgy, and of which the best known was that of the Three Women at
the Tomb, performed on Easter morning. A monk clad in white was to repre-
sent the angel at the tomb, and three monks, vested in copes and holding thur-
ibles, were to approach it ‘step by step as if looking for something’. At the
appropriate moment the angel was to begin singing softly and sweedy the
words ‘quern quaeritis’ (whom do you seek). The stage instructions for this
drama are contained in the English Regularis concordia of c. 970, while the
musical score is in the Winchester Troper (Bodleian Library, Oxford). 5 The
music perhaps came immediately from Corbie, whence Bishop fEthelwold
derived his singing-master for Winchester. It has been plausibly argued drat
replicas of Christ’s Tomb, as shown to pilgrims in Jerusalem, may well have

3 Jantzen (1947), pp. 1 1— 14; Widukind, Res gestae Saxonicae n, 2oandm, 54, pp. 84, 133.

4 Brandt and Eggebrecht (1993b), pp. 221—4.

5 Regularis concordia , c. 5 1 ; Oxford, Bodleian Library, MS 775 , fol. 1 7r.

2l6

HENRY MAYR-HARTING

been connected to these dramas. Bishop Conrad of Constance, for instance,
who died in 975 and who had three times visited Jerusalem, had ‘a sepulchre of
the Lord similar to that at Jerusalem’ made for the church of Saint-Maurice in
his city and adorned with fine goldwork. 6

A feature of early medieval churches which was of great importance but
which is difficult to recover from the surviving evidence was wall-painting.
Wall-paintings depicting the main scenes of Christ’s life must have been
common in Carolingian churches, where they were used to instruct the illiter-
ate, and some remarkable schemes of wall-painting survive from more edu-
cated contexts in the Carolingian period, such as Auxerre, Mustair, Mals and
Brescia, while we know of others, notably at the royal palace of Ingelheim and
the monastery of St Gallen, from written sources. We know extraordinarily
little about how wall-paintings were related to pictures in manuscripts, or for
that matter to each other, at this period. One might be tempted to wonder
whether wall-painting would not have been a vast store of iconography now
lost to us, and whether it is not as likely that wall-paintings were primary
models for book-illuminations rather than the other way round. Final pro-
nouncements seem impossible on these issues, but it has been shown that in
the case of tenth/eleventh-century wall-paintings at Saint-Julien of Tours the
models were ninth-century Turonian book-illustrations, while very few of the
scenes in the large St Gallen Christ cycle of the ninth century are not also
found in Ottoman books, presumably use being made in each case of similar
late antique sources which, practically speaking, could only have been available
in book form. Gauzlin, abbot of Fleury, of whom more will be said later,
obtained a painter from Tours called Odelric to paint the walls of his church
with scenes from die life of St Peter and the Apocalypse. 7 It is hard to see how
their model would not have been books, for there was a rich iconography of
the Apocalypse in Carolingian books, and the St Peter scenes could have been
taken from an illustrated Arator or some such book of illustrated lives of the
Aposdes as Bede had earlier seen. Indeed from the ninth century onwards
there was a brisk business in hagiographic illustration both on walls and in
books, with many of the scenes easily adaptable from one saint to another.
That scenes could have been taken from books and put on walls does not prove
that they were, but it is not easy to cede the primacy to wall-paintings in these
circumstances.

One especially interesting patron of wall-paintings in the tenth century was
Bishop Gebhard of Constance (980-96), in his Eigenkirche, the monastery of
Petershausen. This bishop had the walls of die church covered with pictures,
on the left hand side widi Old Testament subjects, on the right widi New

6 Oudals chalk, Vita Chounradi episcopi Constantiensis , cc. 7, 1 1 .

7 Andrew of Fleury, Vita Gau^lini abbatis Floriacensis monasterii, cc. 62—4.

Artists and patrons

2I 7

Testament. The idea of such sacred parallels was of course an old one, being
found in late antique manuscripts of the New Testament, in the painted
wooden boards which Benedict Biscop acquired for his monastery of
Wearmouth/Jarrow, later in the bronze doors of Bishop Bernward of
Hildesheim, and in revived form in Byzantium after the Iconoclastic
Controversy with their Sacra Parallela. So far, then, the bishop in his murals
was being no more than a professional teacher of his flock. More notable was
the fact that whenever the image of Christ appeared His image was gilded. The
liberal use of gold was an important adjunct of episcopal majesty in this
period; almost all reproductions fail to bring out the way the Benedictional of
dEthelwold of Winchester shimmers with gold. But most remarkable of all was
Gebhard’s abundant use of ‘the Greek colour’, lapis lazuli, for the colouring of
his walls. 8 This was acquired from the Venetians. It has a double significance. It
showed to his people tire far-reaching access of their bishop to luxury goods,
probably at a time when the Ottomans had trade agreements with the
Venetians before the initiative in Venice passed to die Byzantines with Basil II
in 992; and, as a brilliant blue, die colour which in scriptural exegesis always sig-
nified heaven itself, it was an alternative way to tribunes and galleries of con-
veying the idea that a church was a piece of heavenly space.

PRECIOUS OBJECTS

There was a time when all the arts such as diose in gold and silver, enamel,
ivory or embroidery, were known in English parlance as ‘the minor arts’, by
comparison with art in manuscripts, or with painting generally; but happily that
time has passed, and Peter Lasko’s volume of 1971 in the Penguin History of
Art, covering the period 800-1 200 and dealing (admirably) with these arts, was
entidedHrx Sacra, i.e. sacred art. There is a sad refiection even in that tide of the
loss of a whole non-ecclesiastical artistic world. For instance we think of
ivories as covers for liturgical books, or perhaps as pyxes or holy water buckets,
because that is how they have mainly survived. So the following gloss on ‘ivory’
in a late tenth- or early eleventh-century manuscript of Prudentius’
Psychomachia at Cologne may take us aback: ‘ivory’, it reads, ‘is elephant bone
with which die handle of a sword is ornamented’. 9 Such glosses sometimes
refiect only die old books from which they derive, but the comments in this
book are in general far removed from the mere mindless repetition of antique
fiotsam and jetsam. We are familiar widi die art of the goldsmith in various
ecclesiastical forms, but it is refreshing to read that when Otto I wanted to
reward the warrior grandfather of the bishop and chronicler Thietmar of

8 Cames (1966), p. 17. 9 Cologne, Dombibliothek, MS 81, fol 731:.

2 I 8

HENRY MAYR-HARTING

Merseburg for making peace between himself and the archbishop of
Magdeburg, he gave him a golden collar which he wore with pride, to the joy of
his friends and the sadness of his enemies. 10 Would that we could hire a time
machine for a group of art historians to go and study this collar while Count
Henry of Stade was wearing it at a feast! Perhaps diey would find that it had
repousse, foliated scrolls, inhabited by birds and beasts, like the rare but purely
ceremonial Ottoman sword sheath at Essen.

Certain as it is that many finely ornamented books of our period have disap-
peared, the losses in the ecclesiastical world of die ars sacra , not to speak of the
secular equivalent, must have been far greater. C. R. Dodwell has brilliantly
evoked this lost world, and the idea of it as such, in his book on Anglo-Saxon
Art. What he has done for England with the help of literary sources - and one
could use the same method for the continent - can in a way be achieved
through a short cut by using Bernhard Bischoff s edition of early medieval trea-
sure-lists for the empire. 11 Here a breath-taking world of bygone book covers,
reliquaries, chalices, crosses, candlesticks, thuribles, altar frontals, fine linens
and embroidered vestments meets our gaze, and speaks eloquently of die mar-
riage of art and ceremony, of die marriage between the monastic or canonic
life and high liturgical culture, in our period. These lists were drawn up to com-
memorate the munificence of a benefactor to a church or to record the valu-
ables in its treasury or sacristy at the moment when a new custodian or provost
took office.

Did they fantasise, imagining treasures sometimes on a Beowulfian scale
which never in fact existed? Occasionally some part of a treasury survives, as at
Hildesheim with the wonderful artefacts of Bishop Bernward (993-1022),
including two famous silver candlesticks inscribed with his name, or at
Conques in the Massif Centrale of France, where spectacular reliquaries and
other objects have been jealously guarded by the local community down the
ages. Such survivals, and the resplendent book covers which have come down
to us, show that we need not think of these lists as fantasies. But what I wish
particularly to consider is the Ottoman treasure at Essen. Essen was a royal
nunnery and from the time of its Abbess Matilda (971-101 1), grand-daughter
of Otto I, there survive in the first place three splendid golden processional
crosses set with all manner of precious gems and first-rate enamels. They are
likely to be Cologne work. The gold figure of Christ crucified in the most
famous of these has something of the feeling of the limewood Gero crucifix
(probably 969-76) in Cologne cathedral, and the best cut stones, two amethysts
of different shades and a garnet, are reserved for die cross in his nimbus. At its
base is an exquisite enamel, depicting Abbess Matilda and her brother Otto,

10 Thietmar, Chronicon n, 29, p. 78. 11 Bischoff (1967b).

Artists and patrons

219

duke of Suabia, ceremonially presenting the cross itself to the abbey. Another
similar cross has at its foot a small enamel of Matilda, dressed now in pure
white and kneeling at the feet of a hierarchically seated Virgin Mary and Child.
Among the precious stones in this cross are a sardonyx engraved with a fisher-
man, an antique cameo with female bust, and an amber carving of a lion at rest
immediately under the feet of Christ. The third processional cross, from the
same period, has particularly profuse filigree work and enamels of the crucifix-
ion scene and the evangelical symbols.

What are these crosses actually about? First of all they are about religious
ritual and symbolism; manuscript illustrations of the time showing ecclesiasti-
cal ceremonies are full of processional crosses. A nice example of symbolism
is the amber lion, for the Physiologus saw die lion, which supposedly slept with
its eyes open, as typifying Christ crucified, asleep so to speak in his humanity
but wakeful in his divinity. 12 The crosses are, however, just as much about the
authority of the abbess. Matilda was a Liudolfing, but we may be sure that
under her were many other high-born nuns, and we know from die docu-
mented experiences of St Radegund and St Leoba how searchingly the author-
ity of an abbess (as of an abbot) could be put to the test in the early middle
ages. Unexpectedly small as the enamels of Matilda are, once the eye has
caught them, they both make an impact at a distance of as much as twenty feet.
The Matilda/Otto cross must fall more or less widiin die first decade of
Matilda’s forty-year period of office, since Duke Otto died in 982, in other
words during her least secure time, before she could draw on that natural
respect accorded in the tenth century to long life (one need only consider the
analogy of Otto I’s reign and the chronology of rebellions against him). We
know diat tenth-century nuns could be hard-bitten people, and whedier a few
small enamels by themselves could cause many of them to quake at the knees
when they beheld their abbess may be doubted. The important consideration,
however, is what effect this kind of art could have had on die self-confidence
of die actual persons who ruled, on dieir sense of the canonisation of their
own authority. Moreover, coundess medieval hagiographies and histories of
religious houses show us the importance of architectural beautification and
material enrichment for sustaining the authority of an ecclesiastic, and Matilda
herself embarked also on a new campaign of building at Essen which pro-
duced another interesting west work of which a partial impression can still be
obtained. In that west work stands now a remarkable and huge bronze candela-
brum of seven branches with great ornamental knops, while die famous free-
standing gilded Madonna and Child can be seen in another part of the church,
both of them from Matilda’s time.

12 SeeWessel (1966), p. 27.

220

HENRY MAYR-HARTING

There are other fine objects in the Essen treasury from the later period of
Abbess Theophanu (1039-58), grand-daughter of the empress of that name.
One other work from Matilda’s time, however, should not pass without
comment, namely the dazzling gold and jewelled crown, edged with continu-
ous rows of pearls, which may reasonably though not certainly be associated
with her young cousin, Otto III. Here a large red stone is engraved with a
crowned head, while below die fleur-de-lys-type cross which one supposes
represented die front of the crown, in a raised golden setting, is a huge sap-
phire. In the famous representation of Otto III in the Aachen Gospels, the veil
or scroll held by die evangelist symbols not only ‘clothes die heart’ of the
emperor with die Gospels, but also in some way distinguishes a heavenly
sphere of God-given rule from an earthly; in another way this symbol of the
heavenly sphere (as a sapphire was always taken to be), placed so prominently
on this crown, seems to make the same point. 13

The greatest of all goldsmidis’ workshops in our period were those of
Archbishop Egbert of Trier (977—93), which carried out commissions for
other churches, such as the golden book cover (still surviving in Nuremberg)
which the Empress Theophanu ordered for the monastery of Echternach, or
the enamelling with which they adorned a golden cross from Rlieims (as is evi-
denced by a letter of Gerbert of Aurillac). 14 We know from the work of
Hiltrud Westermann-Angerhausen that Egbert did not bring these workshops
into being, for diey were in Trier before him, 15 but he did use them richly and
imaginatively to propagate an image of the majesty of his church as well as a
sense of Christian mission. 16 Amongst the indated claims to antiquity which
the great Ottoman churches made in their rivalries with each other, Trier
claimed to have been founded by none other than St Peter, and to validate its
claim it preserved the relic of St Peter’s staff, which Egbert had encased in a
gold and bejewelled container (still surviving in the Limburg cadiedral treas-
ure), with enamelled representations of the earliest bishops of Trier, the first
diree, Eucharius, Valerius and Maternus, reputed followers of Peter himself.
As if that were not apostolic support enough, Egbert also had St Andrew’s
foot, a great relic of his church, placed in a priceless container, a chest with the
finest enamels set into it, topped by a golden foot (to be seen in the Trier cadie-
dral treasury). Egbert was also a great patron of illustrated books, in two of
which, a book of Gospel readings known as the Codex Egbert! (at Trier) and a
psalter (at Cividale del Friuli), he was himself depicted on the frontispiece in
hieratic posture, staring straight ahead as if seated in a world divorced from the
earthly one. In these and other books which were written under Egbert,

13 Aachen, Dom, Schatzkammer. 14 Gerbert of Aurillac, ep. 106.

15 Westermann-Angerhausen (1987); (1990), esp. p. 20 .

16 Westermann-Angerhausen (1983); (1973), pp. 66—72.

Artists and patrons

221

Westermann-Angerhausen has shown that there are depicted capitals with
masques and other ornaments, which would have derived from the capitals of
Bishop Nicetus’ sixth-century cathedral, covered over by Egbert’s own build-
ing works on the cathedral. Egbert had twice visited Italy in the entourage of
Otto I and Otto II, but Trier itself had an imperial and historic past, if not quite
so historic as its church claimed at this time. It had Roman buildings and other
remains (a gold coin of Constantine is set into the St Andrew reliquary), and it
had Carolingian books, derived from Tours, which were demonstrably the
source of many of the ornamental motifs used in the goldsmiths’ workshops. 17

As patrons of art Archbishop Egbert and Abbess Matilda may have had at
least one point in common, desire to bolster a vulnerable authority. It is easy to
think of Egbert’s art as expressing the pinnacle of might reached by one of the
dominant churches of the tenth century; but it was probably otherwise, for in
the ecclesiastical power game of that time, Trier under Egbert appears to have
been losing rather than gaining influence. Averil Cameron has shown that sim-
ilarly in ninth- and tenth-century Byzantium new rituals (and new art) are as
much a response to political pressure as an articulation of effortless superiority,
that the divine and earthly harmony of imperial art and ceremony could gloss
over a much tenser reality. 18 Authority is again at issue in a most remarkable
phenomenon of the goldsmith’s art in southern France during our period, the
statue-reliquary, of which the extraordinary example of St Faith at Conques
survives, a statue of gold (over wood) studded with jewels, and representing
the saint seated in a hieratic posture like an oriental potentate. Public authority
in that region had given way to castellans like the counts of Rouergue exercis-
ing power from local castles, and the monastery of Conques was anxious to
stress, in this image, that its authority over its lands and men was embodied in a
patron saint well capable of giving predators nightmares and worse. All the
same, the northern French rationalist Bernard of Angers, a pupil of the great
master Bishop Fulbert of Chartres, disapproved of such things. They were
common in the Auvergne, he observed, but were a form of idolatry of which
he doubted that Jupiter and Mars would think themselves unworthy. 19

A great collector of objets d’art, as well as patron of artists, was Gauzlin,
abbot of Fleury (1004-30) and half-brother of Robert the Pious; doubtless
this relationship to the king of France put him in a useful position to accumu-
late goodwill offerings, and his monastic biographer, Andrew, was nothing
loath to detail them. Bishop Bernard of Cahors, an alumnus of the Fleury
school, gave him a golden altar frontal and some finely embroidered altar
cloths; and Arnold, count of Gascony, sent him thirteen silver vasa and two
pounds of Arabic metal’ as well as some oriental silks. The abbot himself had a

17 Westermann-Angerhausen (1987). 18 Cameron (1987).

19 Uber miraculorum sanctae Fidis 1, 13.

222

HENRY MAYR-HARTING

lectern of ‘Spanish metal’ made, as well as using Spanish copper plaques to
enclose die choir at Fleury, while the precentor Helgaud, biographer of King
Robert as well as craftsman, made a precentor’s baton widi a handle of crystal,
sparkling widi precious stones. 20 We are seeing in all this, amongst other things,
the build up of a luxury traffic between Spain and die region of the Meuse,
connected by the rivers Rhone and Saone and spinning off into other parts of
France. The general importance of this great river trade route was long ago
established by Maurice Lombard, and in 1 9 5 o J. M. Lacarra published a remark-
able customs document issued by King Sancho of Navarre in the 1 060s, detail-
ing commodities which passed through Jaca at the foot of the Somport Pass in
the Pyrenees. 21 ITere we can see dyed Flemish cloths and various forms of
weaponry coming down from the north, while Constantinopolitan textiles,
Castilian horses and Spanish gold are coming up from die south. Awareness of
Spanish wealdi helped to draw French knights into an involvement with the
reconquista in the eleventh century, and the Cordovan as well as Greek textiles
and silks witnessed in King Sancho’s document contained, we can be sure,
those zoomorphic stylisations and geometric designs which held such pro-
found inspiration for the Romanesque sculptors who would translate them
into stone capitals.

One cannot ignore the function of fine objects for a political world whose
relationships were still sustained as much by gift exchange as by legal contract.
The Emperor Otto III, who sought to run his relationship with Doge Peter II
Orseolo of Venice (991— 1009) along the Byzantine lines of expressing his
superiority through the godfatherly status, was offered rich gifts by the Doge
when he visited Venice in 1000. He did not accept them all, for fear of looking
as if he had visited Venice not purely out of love for his godson (who would
have thought of any other motive?), and finally left with only an ivory chair, a
silver goblet and a jug with rare ornament. 22 Probably it was important for him,
as the greater ruler, to accept gifts only of lesser value than he would give, and
he subsequently sent the Doge fine works of gold from Pavia and Ravenna.
Again, the courtiers of tire Emperor Henry II loved to visit Magdeburg
because they were always rewarded with splendid gifts from the archbishop,
who could tap into an important trade of oriental luxury goods passing down
the River Elbe.

BO OK- ILLUMINATI ON

Whatever the relatively low importance of book-illumination amongst die arts
to contemporaries of our period, it speaks to us across the centuries, partly

20 Andrew of Fleury, V'ita Gau^lini abbatis Floriacensis ?nonasterii, cc. 38, 39,65,47.

21 Lacarra (1950), text pp. 19— 20. 22 Uhlirz (1964).

Artists and patrons 223

because of its high rate of survival compared with artefacts in other forms of
art, and pardy because it is certainly no less effective a medium to express the
ideas, attitudes and aesthetic of die age. The principal problem in making an
even study of the whole west European world of book art is that the books of
the German, or Ottoman, empire bestride that world like a colossus. However
much we may give parity to French/west Frankish or Italian metalwork or
ivories, when it comes to book-illumination, tire Ottoman art runs away with
the prize.

One may say this while acknowledging that there were undoubted master-
pieces of book art in other regions. In England the Benedictional of St
dEthelwold, the chef d’ceuvre of the Winchester School (c. 970-80), reigns
supreme, with its early openness to Byzantine influence, and its majestic Christ
scenes and saints’ images framed with ebullient foliage ornament derived from
Carolingian manuscripts of Rheims. There is also a wealth of other English
manuscripts, such as tire Bury St Edmunds Psalter (c. 1020) with its brilliant
drawings in the margins, while the debt to Anglo-Saxon art of the great art
patron and artist Abbot Otbert of Saint-Bertin (probably 989-1007) in
Flanders, particularly in his Gospel Book and Psalter now at Boulogne, is gen-
erally acknowledged. English book art was very different from Ottoman.
Expressed briefly, one might say that whereas Ottoman illumination derives its
character from work in gold and enamel, English art is more linear (to take up a
theme expounded by Nikolaus Pevsner in his The Englishness of English Art)? 7,
more draughtsman^, more closely related to the world of its great calligra-
phers, though great calligraphers worked on Ottoman books too. The impulses
of patronage, however, from kingly rule and the episcopal Tremendum which
was so important a means of sustaining that rule, were similar in both societies;
dEthelwold, bishop of Winchester (964-84), is comparable to Egbert, arch-
bishop of Trier (977-93), in his closeness to a royal court and in the projection
of his own mighty image through art. The vital difference — and it is a point
which always has to be borne in mind when considering art patronage in this
period - is that while dEthelwold’s art reflects a waxing of Winchester within
the English church, Egbert’s would appear more as a response to die waning of
Trier within the imperial church. 24 dEthelwold rides on the crest of a wave;
Egbert faces the pressure of a stormy sea.

The finest French book art of tire tenth and early eleventh centuries drew its
inspiration first and foremost from Carolingian and pre-Carolingian traditions
of that region. The so-called First Bible of St Martial, Limoges ( c . 1000, BN
lat. 5), has a wonderful series of ornamented initial letters with very lively zoo-
morphic and plant decoration, which takes one back to the great bibles of

23 Pevsner (1956), ch 5. 24 Mayr-Harting (1991b), pp. 82— 8.

224

HENRY MAYR-HARTING

ninth-century Tours, and, in their animal forms, to works such as the eighth-
century Gellone Sacramentary. We are dealing here with first-rate artists, but
ones who appear indifferent to die figural art of the Ottoman books, or indeed
of the Turonian bibles. The concern of diese artists was not with projecting
the image of an all but non-existent west Frankish royal power or of an episco-
pate which had little existence independent of the aristocratic power structure;
it was with embellishing the studies and liturgy of an important monastery far
from the islet of effective Capetian rule.

N either Spain nor Italy is lacking in notable works of book art in our period.
In Spain we have above all the series of illustrations to Beatus’ Commentary on
the Apocalypse. If one wishes to see in something of their purity die late
antique traditions of Apocalypse illustration, which were transformed in style
by the Mozarabic modifications, one should consider (as Florentine Miitherich
has pointed out) the Beatus commissioned by Abbot Gregory of Saint-Sever
(1028-73), f° r which the artist used a Spanish model. 25 In south Italy there is
the rich series of Exultet Rolls, and, in the eleventh century, the rise of narra-
tive illustration connected to the life of St Benedict, at Monte Cassino; in the
north there were interesting provincial schools like that of Bishop Warmund
of Ivrea (c. 969-101 1), another bishop who liked to have himself depicted in
books, but who in real politics was (as a supporter of the Ottomans) under
great pressure from Arduin of Ivrea. The illuminated books of Warmund,
principally his Sacramentary and Psalter, both at Ivrea to this day (Codd. 8 5 and
86), show a variety of iconographic influences from Carolingian and Ottoman
art, and no doubt from earlier Italian books which also influenced these. The
Psalter shows the liking for monumental standing figures evidenced also in the
Prayer Book of Archbishop Arnulf of Milan (r.iooo, BL, MS Egerton 3763),
and in early Italian wall-paintings, as Hans Belting has shown. 26 The Warmund
style, however, is a world away from anything Carolingian or Ottoman,
showing how little political influence carries with it the assumption of accom-
panying artistic influence. The draughtsmanship is clumsy, but for all that, its
potentiality for great liveliness is realised in a series of feverish illustrations to
the ordo in agendis mortuomm in tire Sacramentary, showing the death and burial
of a man while his grieving wife or mother becomes more and more dis-
traught, until she has to be restrained at the graveside. As to the principal
colour tones of blue, green, pink and yellow in the Sacramentary, they have
little to do with the work of goldsmiths and enamel workers, but rather more,
perhaps, with that of muralists who were so important in north Italian art at
this time. One may confirm this from the nearby contemporary wall-paintings
in the cathedral baptistery at Novara.

25 Miitherich (1973), pp. 195-6. 26 Belting (1967).

Artists and patrons z z 5

Because of the supremacy of Ottoman book-illumination in our period,
however, we shall make our points about patrons and artists largely through it.
First, we have to ask how Ottoman book-illumination began, after the long
hiatus (for the most part) from the late ninth century to the 960s, generally pre-
sumed to be caused by external threats and unsteady politics. My answer would
be that Otto I and his court played no small part as a stimulus, but this must
remain in the nature of an argument or a hypothesis, for it cannot be proved.
The Ottomans never had any court school of illumination as the Carolingians
had, where production of de luxe manuscripts, especially of the Gospels, was
directly under their control. Under Otto III and Henry II, and also under
Henry III, certain monasteries, such as Reichenau, St Emmeram of
Regensburg, and Echternach, worked for the ruler, but in the cases of Otto I
and Otto II we cannot even name any surviving manuscript which was cer-
tainly produced for either of them. The case for Otto I’s stimulus, therefore, is
based on a number of convergent indicators. His general interest in books is
explicitly testified, not in the early part of his long reign when he gave a very
exiguously ornamented Gospel Book to his brother-in-law, fEthelstan of
Wessex, but in the later part. Widukind says, ‘after the death of Queen Edith
[946], whereas previously he knew nothing of letters, he learned them to such
an extent that he could read and understand books fully’. 27 Culturally Otto I’s
horizons broadened manifestly in the latter half of his reign. In 968 he finally
established the archbishopric of Magdeburg as a lynchpin of his ecclesiastical
organisation to the Slav east, together with several suffragan sees and depen-
dent monasteries. In his foundation document he stressed his own initiative,
and indeed this had been a central project of his since 955. Suddenly a huge
new need arose for fine liturgical books as well as library books; Otto I realised
this, for Thietmar of Merseburg says that he endowed Magdeburg generously
‘in estates, books, and other royal splendour’ (suggesting ornamented books). 28

There seems little doubt now that perhaps the two greatest works in the
initial Ottoman revival of book-illumination, tire Gero Codex (Darmstadt, MS
1 948) and die Codex Wittekindeus (Berlin, Staatsbibliothek, MS theol. lat. fol.
1), were made respectively at die monasteries of Reichenau and Fulda. Both
these monasteries had particularly close connections with Otto I and his court;
the emperor is known to have visited Reichenau in 965 and again in 972, the
very period when the Gero Codex would have been in the making. Moreover
the Gero Codex, which is a book of Gospel pericopes, has a liturgical calendar
in which the only non-New Testament based feasts are those of St Laurence
and the Maccabees. These two feasts were amongst die normal celebrations
of the Roman calendar in the tenth century, but singled out in this way they

27 Widukind, Res gestae Saxonicae n, 36, p. 96. 28 Thietmar, Chronicon n, 30, p. 76.

226

HENRY MAYR-HARTING

represent a virtual hall-mark of Otto I because both bear significantly on his
victory over the Hungarians at die Lechfeld in 9 5 5 . He spent the tenth anniver-
sary of this victory, the feast of St Laurence 965, at Merseburg, where he had
vowed on the batdefield to establish a church in honour of drat saint whose
feast it was. The liturgy of books and the liturgy of public life were not two
separate issues in his time. The Maccabees’ resistance to the Seleucids was seen
as a veritable biblical type of the Ottomans’ resistance to the invading
Hungarians. The Gero who commissioned the book and is shown receiving it
cannot be identified with certainty, but there are several good reasons for
regarding that Gero who was archbishop of Cologne from 969 to 976 as the
likeliest candidate, and he had been a court chaplain of Otto I before his eleva-
tion. As to the Codex Wittekindeus, a book of the four Gospels, its earliest
known provenance was the monastery of Enger, which was granted by Otto I
to the archbishopric of Magdeburg in 968.

The greatest book painter of the early period of Ottoman artistic efflores-
cence was die Gregory Master; he was an expert calligrapher and furthermore
no other artist shows such mastery of how to handle his late antique proto-
types, their modelling and perspectives. We cannot name him; indeed it is a sad
fact that we cannot name a single Ottoman book painter in relation to any par-
ticular book; but we can trace him through his work in a period of activity
which spanned die last three decades of the tenth century. Art historians have
given him his tide from a double page of miniatures depicting Pope Gregory
the Great dictating his Dialogues (Trier, MS 1 71/1 62 6) and the Emperor Otto II
seated in majesty surrounded by personifications of imperial provinces
(Chantilly, MS 146). He appears to have been based at Trier in Archbishop
Egbert’s time (977-93), but he also worked with and for the churches of
Lorsch, Reichenau and probably Fulda. He was very peripatetic, which might
suggest that he was a layman but was by no means incompatible with his being
a monk in those times of so many connected monasteries with dieir confrater-
nity arrangements. Now if Hartmut Hoffmann is right that he was the artist of
die Marriage Roll of the Empress Theophanu (972), a magnificent document
written in gold letters on purple grounds with vivid drawings of lions and
griffins, and the case is a persuasive one, 29 then diis becomes his first known
work, and it was produced under the patronage of the Ottoman court in the
reign of Otto I. Hoffmann has also established that the script of this document
was Fulda, and so it was produced in conjunction with Fulda, around the
period of the Codex Wittekindeus. And by 972 Egbert, who would become the
Gregory Master’s principal patron, was a court chaplain. Hoffmann himself
prefers to regard this Marriage Document as a feature of the new culture of

29 Hoffmann (1986), pp. 103— 16.

Artists and patrons 227

Otto II, co-ruler with his father from 96 1 , rather than of the aurea mediocritas of
Otto I. It makes little difference to die main argument — die likely court stimu-
lus of early Ottoman book art, even though there was no court school.

The high point of Ottoman book-illumination comes with a series of books
made at Reichenau or Regensburg for Otto III (983-1002) and Henry II
(1002-24), such as the Aachen Gospels (Aachen Minster), the Gospel Book of
Otto III at Munich (Clm 445 3), the Bamberg Apocalypse (Bamberg, Bibl. 140),
the Pericopes Book of Henry II (Clm 4452) (all Reichenau) and the
Regensburg Sacramentary of Henry II (Clm 4456). These are amongst the
summits of western civilisation. They contain images of the rulers which give
the royal/imperial ideology a very high pitch, not to speak of their superlative
series of New Testament scenes. Of the ruler image in the Aachen Gospels,
for instance, Hagen Keller has observed that, whereas Carolingian kings are
depicted as in this life and as interacting presidents of their courts, Otto III is
removed from his sub-kings and courtiers and sits in a ‘super-earthly sphere’,
stiff and frontal in posture like a Christ in Majesty. 30 That is typical for the
Ottomans.

Given drat Reichenau and Regensburg were not court schools but rather
monasteries which undertook work for the court, it is reasonable to pose the
question whether these ruler images stem from die court ideology, whether
they are painted on the instructions of die court patrons so to speak, or
whether the monasteries themselves actually formulated this ideology through
art in order to win court favour for themselves. These are, however, stark alter-
natives; they suggest too low an idea of die cultural integration of the great
imperial monasteries with the court itself. We have to remember that Ottoman
kingship was itinerant, a fact which would have brought the kings to such mon-
asteries more often that we can now tell from the surviving evidence.
Moreover their abbots were often close friends, or familiares, of the rulers; in an
itinerant kingship the circle of familiares is not confined to those ‘at court’.
Abbot Alawich II of Reichenau (997-1000) was on friendly enough terms with
Otto III to join him at Rome in 998 and to be made bishop of Strasbourg by
him two years later. Henry II knew personally not only Abbot Berno of
Reichenau (1008—48) but also others of the monks in the monastery, while he
had himself been educated at St Emmeram of Regensburg, as had his earliest
principal adviser, Tagino, archbishop of Magdeburg (1004-1 2).

We do not have any evidence to know how court patrons actually dealt with
monastic artists in the case of Ottoman books, but in so far as we can make
deductions, these have to allow for a positive court input to explain the ruler
imagery and other ruler-related imagery in diem. For instance, the Otto III

30 Keller (1985), pp. 302—5.

228

HENRY MAYR-HARTING

image of the Aachen Gospels, which most commentators would now date to
after his imperial coronation of 996, appears to draw on the inspiration of
tenth-century Byzantine ivories of the Ascension, depicting Christ not stand-
ing in the usual western fashion, but seated on an orb as he ascends to heaven.
Reichenau had been an eager recipient of Byzantine culture throughout the
tenth century; however, this is not something which sets it over against Otto
Ill’s court, but something which tire two institutions share. More particularly, it
is hardly likely that Reichenau was responsible for fixing Ascension Day, with
all its connotations of Christ ideology and apotheosis, as the feast on which
Otto Ill’s imperial coronation was set in 996. Equally the Rome emphasis in
the ruler imagery of the Munich Gospel Book could not possibly be explained
without reference to the influence of the court chaplain, Leo of Vercelli, on
Otto III in this respect. The rare splendour for this period of the depiction of
St John the Baptist’s Nativity in the Pericopes Book of Henry II, and its hier-
atic character, may have been die idea of the Reichenau artist, but if so, it
cannot have been conceived without a good knowledge of ideas already exist-
ing in Henry II’s head. For he had celebrated the feast of the Baptist’s Nativity
at Reichenau itself in 1002 during his Umritt , that is, when he travelled around
his kingdom to gain acceptance for his kingship by public ritual acclamation
after an intense struggle earlier in the year. It was certainly not Reichenau which
was responsible for the subsequent emphasis on the Umritt as a validation of
Henry’s kingship. One could say much more about die correspondence
between court thinking and ruler imagery if space permitted.

The question whether it was patron who specified the ruler images or artist
who suggested them is dierefore to some extent an unreal one. When Rubens
painted his great cycle of pictures glorifying the Regency of Marie de Medici,
die latter’s conception of her political persona and aims are not the less domi-
nant in Rubens’ scenes because many of their subjects derived from his own
suggestions. Rubens was a learned man and so were many Ottoman artists. We
may not be able to name any artist in connection with a particular work, but we
know something about artists genetically. The scheme of illustration for the
Uta Codex of Regensburg, highly dieological in content, was devised by a
monk called Hartwic, who had studied under die learned Fulbert of Chartres. 31
A Trier artist called Benna, painting at Wilton in the 980s, was not only
renowned for his art but also respected for his learning. At Fulda, whose main
business in book-illumination appears to have been the production of mass-
books, not least for export, we know, from the monastery’s records of deaths,
of a person called Ruotbraht, subdeacon, monk and painter (pictor could mean
a wall painter, or book painter, or most likely both), who died in 977. A subdea-

31 Bischoff (1967a).

Artists and patrons

229

con, if only in his twenties, must have had a certain degree of learning as well as
clearly defined liturgical responsibilities within the Mass. Whichever way we
look at it, therefore, the most satisfactory idea to have in mind is neither that of
the rigid orders of a patron nor that of the surprise packet of a clever artist, but
of an interaction between die requirements of patron and their creative realisa-
tion by artist, through intelligent dialogue.

This area of relations between patron and artist is where an important point
established by Hartmut Hoffmann may fit in. In several Ottoman manuscripts
we find a picture of a cleric proferring the book to its ultimate earthly recipient
(I am not speaking here of die traditio to a saint), such as Liuthar to Otto III in
the Aachen Gospels, or Ruodpreht to Archbishop Egbert in die Egbert Psalter
at Cividale (MS 136), or the two Reichenau monks Kerald and Heribert in the
Codex Egberti (Trier, MS 24). Hoffmann has shown that such a figure would
perhaps never have been the artist, and that only in rare cases can he be said to
be the scribe, as with Eburnant in the Hornbach Sacramentary of Reichenau.
Stifter, or donors, is the term he uses for these clerics. 32 That does not necessar-
ily mean that they paid for the materials and work of the manuscript; though it
could mean that, even if such a one were a monk, for monks often had rich
families. Liuthar, Hoffmann says, could have been the scribe, or he could have
been die current leader of the Reichenau scriptorium, but what matters is that
he acts here as a respected representative of his community. Ruodpreht, of the
Egbert Psalter, is an even more interesting case. If he was a scribe, why should
he be singled out amongst the several Reichenau scribes whom Hoffmann
shows to have participated in the manuscript? Indeed, he need have had
nothing to do with Reichenau, and was probably a monk or abbot of Egbert’s
circle, the Stifter. Now when one studies the Codex Egberti and the Egbert
Psalter as a whole (as I have done elsewhere), it is clear that they are deeply shot
through with Archbishop Egbert’s own concerns and preoccupations. Their
mode and matter can in no way have been left to the unaided discretion of
Reichenau. Egbert himself had probably visited Reichenau at least once during
his archiepiscopate, on his way back from Italy in 983 . But Egbert’s protege, the
Gregory Master, himself painted the first illustrations in the Codex Egberti,
and should he not be seen, together perhaps with the Reichenau monks Kerald
and Heribert, as the ideal mediator between patron and scriptorium? Likewise,
perhaps, Ruodpreht in the case of the Egbert Psalter?

CONCLUSION

As we contemplate Ottoman art we are drawn back ever and anon to liturgy, to
art as a means of ritualising religious experience and political power relations.

32 Hoffmann (1986), pp. 42— 7.

230

HENRY MAYR-HARTING

Rulers, especially Henry II, appear constantly as if they were the central actors
in church services, and several sacramentaries, or mass-books, of Henry II’s
time carry in their calendars the day of his kingly consecration, his dies ordina-
tionis. It is vital, however, not to treat Ottoman art as if it were all ideology, even
religious ideology, that is, as if its sole function was to be an instrument in the
Ottoman power game. In any power game religious art would be a worthless
instrument unless it could appeal to a body of believers whose own religious
experience was at least in some degree independent of political motivations.
That is why it is important to study the religious culture of the great centres of
Ottoman artistic production without seeing politics round every corner, and
the religious culture of their patrons. For example, let us by all means remem-
ber that in the Munich Gospel Book of Otto III the Christ scenes project an
image of a Christ-Emperor, who thereby in some sort canonises the authority
of die earthly emperor; but let us not overlook, when we contemplate, say, the
poignant scene of the Repentant Mary Magdalene in that same book, that Otto
III owned a prayer book, one of whose prayers, headed ‘whoever prays this
prayer shall not feel die torments of hell in eternity’, says, ‘be mild to me as you
were to Mary the whore, and fill my eyes widi tears as you filled hers when she
washed your feet and wiped them with her hair ’. 33

33 Pommersfelden, Schloss, MS 347, fols 3 ir— 34V.

PART II

POST- CARO LIN GIAN EUROPE

CHAPTER 9

THE OTTONIANS AS KINGS
AND EMPERORS

Eckhard Miiller-Mertens

THE POST-900 POLITICAL CRISIS: THE DISSOLUTION OF THE
FRANKISH KINGDOM RULED BY ARNULF AND THE DEVELOPMENT
OF REGIONAL AND ETHNIC PRINCIPALITIES

The Emperor Arnulf died in Regensburg on 8 December 899. The illegitimate
son of King Carloman of Bavaria and Italy had brought about the fall of
Charles the Fatin November 887, which had led to his own election by the east
Frankish magnates and to the election of non-Carolingian rulers in other parts
of the Carolingian empire. Charles the Fat had been able to reunite the
Carolingian kingdoms and, apart from Provence, had exercised direct rule over
all of them. Unlike Charles, who had accepted the west Frankish crown offered
him in 885, Arnulf of Carinthia rejected a corresponding offer from the west
Frankish magnates. This incident, whose significance, especially for die devel-
opment of a German kingdom, has been much discussed, did not mean that
Arnulf of Carinthia wished to confine his rule to east Francia, Francia orientalis.
Fie established a relationship of feudal overlordship, or at any rate allowed one
to be established. The other rulers elected in 888, Odo of west Francia, Rudolf
of upper Burgundy, and Berengar of Italy, as well as, later, Louis of Provence,
acknowledged his overlordship. Fie sent Odo a crown, with which Odo had
himself crowned a second time in Rheims. After Charles the Simple had been
set up as king in west Francia in 894 he too submitted to Arnulf, who acted as
mediator in the dispute between Odo and Charles over die west Frankish
throne. Arnulf disputed the claim by Rudolf of upper Burgundy to rule over
the whole of die former kingdom of Lodiar II; and when Wido of Spoleto
challenged his overlordship by having himself made emperor, Arnulf inter-
vened in Italy and acted directly as Italian king.

Arnulf, as king over other kings, exercised an imperial kingship. It was in
keeping with this when in 8 9 5 he transferred die kingdom of Lodiaringia to his
son Zwentibald. The latter’s newly independent position, within die ambit of
imperial kingship, was intended to act as a check on the aspirations of Rudolf I.

2 3 3

North Sea

« 6'
,»\b

\ O^MARCH OFTHE f
I o-* BILLUNGS I

Hamburgh

jVerden

SAXONY

NORTH

MARCH

, '■‘ss/ ,• , .Minaen i? -il ^

I Utrecht Osna brOck Hildesh e iJj|

33^' Munstert Gandersheim^ Jg||§

r^KN.jrnegen^ Psderborg' 5 orvey Qu^Iinby |lf

Poles

Gandereheim^^g

MARCH>OF LAUSITZ

LOWER/

WESTPHALIA

/'/ ' , . v ^-Or" MARCH OF

\ Memleben*-% 1 ^^|Vierset)Ufg \ MEISSEN

7 Cologne j|
^Aachen

, Naumbui

S THURINGIA

Hersfefd \ ,■>
Fulda 1 *-' ^

( 7 'M ^

— ? \ * x ' i

L O T H A R I N G l A'

Verdun J

! I N G l A FRANCONIA /

-'■ > / (HESSE)

/ l "'/ Vi Mainz Frankfurt

o £\ f\ ■

' ^ty/lngelheiA W 7 V l ||Bamberg

Trier""^ t 'A... w// " 3

f l ^tWorms Wurzburg ^

U P PrE R (<\ SDever^^ - ^ V n

Boh

e m i an s

KINGDOM

OF FRANCE

> PrE R Cs % s

/Metz \

L by H A R | N G I A/~

A /TouV~\ j ‘Wi.

V / l Strasbourg/

[/ COUNTY

y of

BURGUNDY

s' 1 , «

— ^ N , \

Eichstadtt

Augsburg U >

Regensburg

PassairXp^

MARCH OF

SWABIA

• Freising

BAVARIA

SaJzburgt

> X

i Constance^
\ StGallen*

MARCH OF

KING D/u M OF

BURGU N DY

KINGDOM

CAR I NTH I A

MARCH OF

V. MARCH ' w
JOF KRAIN"

/'"X

'MARCH OF ^

Approximate boundary of German
kingdom c. 1000

Regional boundaries

Main concentration of royal lands
t Archbishoprics t Bishoprics

236 ECKHARD MULLER-MERTENS

This policy was a successful one until the point at which Arnulf was attacked
by an illness, shortly after his imperial coronation in February 896, which in the
end made him incapable of ruling. He exercised his kingship itinerantly: its
core regions were Bavaria, with its centre Regensburg, and die lands around the
confluence of the Rhine and the Main, with their centres Frankfurt, Tribur and
Worms as the principal locations for meetings with die magnates. 1 For jour-
neys between these core regions he preferred the Main valley area of
Franconia, widi the royal palace of Forchheim as the place of choice for
assemblies. Arnulf visited Suabia and Lotharingia only occasionally, and
Saxony only once, in the course of a campaign against the Abodrites. Yet he
was nevertheless able to exercise influence in diese provinces: indeed, Suabian,
Lotharingian and Saxon churches and magnates received significandy more
royal diplomata dian did their counterparts in Bavaria and the Rhine-Main
area. Arnulf drew his counsellors from the high nobility of all regions of his
kingdom, including members of families which were later to produce dukes:
Conradines, Luitpoldings and Liudolfings. The bishops Hatto of Mainz,
Solomon III of Constance, Waldo of Freising and Adalbero of Augsburg
played a significant role at the court as well as acting as a link between Suabia
and die king. There were certain differences between Arnulf’s treatments of
the two core regions: die lands around Rhine and Main, and Bavaria. The
Rhine-Main area was of greater importance for assemblies dealing with regnal
affairs, for synods and for meetings widi the magnates from other parts of the
kingdom. Bavaria was less significant as a centre of integration for the
kingdom: it had more die role of a base domain for Arnulf’s kingship, with sig-
nificant direct seigneurial exploitation and intensive contacts between Arnulf
and the Bavarian magnates. These can be seen in die large number of individ-
ual Bavarian recipients of diplomata, both ecclesiastical and secular, who did
not receive their diplomata at assemblies, unlike the practice in the Rhine— Main
area. The latter was easily the most important central region for the politics of
the kingdom. Bavaria came a poor second here, playing rather the role of a

1 Translators note: In the discussion here and at intervals in what follows it may be helpful to have the
terminology of German medievalists, as developed by Miiller-Mertens, Moraw and others,
explained. In this the spatial divisions of the kingdom are conceived of as having different aspects.
As Zonen (zones, which may be ‘distant’, ‘open’ or ‘close’), one thinks of these regions primarily in
terms of the way in which the elites in politically or geographically determined areas saw themselves
in relation to the ruler and the consequent political opportunities for the ruler there. As Landschaften
(translated here as domains), one thinks of concentration or absence of royal resources in terms of
palaces, fiscal lands and rights within an area (so that a ‘base’ or ‘core’ domain is one with a particu-
larly high concentration of such resources). Finally, as Raume (translated here as regions), one thinks
of the absence or presence of the itinerant ruler himself and of the magnates in attendance on him:
there are ‘central’ regions where the ruler stayed for long periods, ‘transit’ regions which he visited
not infrequently but usually on the move between ‘central’ regions, and other regions where the ruler
was rarely found. See now Bernhardt (199$), especially pp. 45—70.

The Ottomans as kings and emperors 237

basis for Arnulf’s own power; he had made his bid for the kingship as margrave
of Carinthia with the backing of an army of Bavarians and Slavs. All this dem-
onstrates the east Frankish and Carolingian structure of Arnulf’s kingdom,
though his chancery no longer used the term Francia orientalis for the kingdom:
when either it or the historians of the period used a name at all, it was plain
Francia.

Arnulf of Carinthia left a single legitimate son, who was still a minor. The
magnates of the kingdom were soon agreed on the succession, and had no
qualms about setting up die six-year-old child as king. This was done on 4
February 900 in Forchheim. Louis IV, the Child, received the allegiance of
Zwentibald’s followers, who had defected from him on Arnulf’s death, shortly
after this in Thionville. Arnulf’s realm thus continued to lack a ruler capable of
acting, for the emperor had already lost control of events as a result of his
illness in the final years of his reign. This was to call in question the characteris-
tic elements of Arnulf’s earlier rule: the position of imperial kingship and the
predominance of direct royal rule within east Francia. A loosely organised,
legally undefined regency, which included the Bavarian margrave, Leopold, and
Bishop Adalbero of Augsburg, carried out the government on Louis’ behalf.
The numerous diplomata include an unusually high number of Frankish,
Suabian and Bavarian intervenors. The principal points on the itinerary were,
as they had been under Arnulf, Regensburg and the palaces of Rhenish
Franconia. After 907, however, the regency council withdrew from Bavaria,
which ceased to be a base domain for Louis’ kingship. The council was domi-
nated by magnates whose power-base lay in the Rhine-Main area: Archbishop
Hatto of Mainz and his church, and the Conradines (Conrad die Elder and
later his son Conrad die Younger, and Gebhard). Hatto concerned himself
with the question of the unity of the empire, and may even have contemplated
an emperorship for Louis, but the practical politics pursued by these men were
rooted in regional issues of rank, property and power in Franconia,
Lotharingia and Thuringia. The Conradines were able to establish themselves
in Lotharingia, where Gebhard received ducal office on behalf of the king.
This produced rivalries with the Matfridings; after their defeat, the Reginarids
took over their position and worked against the Conradines and for the defec-
tion of Lotharingia. In Franconia the Conradines pursued a struggle for supre-
macy with the Babenberger; Conrad the Elder was killed in the course of this,
but with the assistance of Hatto of Mainz the Conradines won, and after the
Babenberger Adalbert had been executed the lord and duke of Franconia was
Conrad the Younger, the future King Conrad I.

In the politics of the kingdom it was now die regna, the large-scale political
areas, which dominated the stage. Franconia and Lotharingia were former
royal provinces, whose political organisation stemmed from the Carolingians;

238 ECKHARD MULLER-MERTENS

Suabia, Bavaria and Saxony were ethnicaUy defined regions. It was here that
rival noble kindreds struggled for supremacy and leadership; it was here that
the defence against the Magyars was organised and led; and it was here that the
transformation of ducal power into viceregal or quasi-regal positions took
place. In Suabia the Hunfriding Burchard of Raetia sought supremacy; he was
opposed by Solomon III of Constance and by the Alaholfings Erchanger and
Berthold until his murder in 91 1. In Bavaria and Saxony tire development of
ethnically based dukedoms by the Luitpoldings and the Liudolfings respec-
tively took place without such rivalries. Margrave Luitpold fell in 907 in battle
against the Magyars, and this marked the loss of the Carolingian marches on
the south-eastern frontier. Luitpold’s son Arnulf was able to defeat the
Magyars on several occasions. In Saxony the Liudolfing Otto the Magnificent
was able to extend his hegemony in Saxony to cover Thuringia following
Margrave Burchard of Thuringia’s death fighting the Magyars in 908. Having
achieved his majority, Louis himself led an army against the Magyars in 9 1 o. He
was defeated near Augsburg; the numerous dead included tire Lotharingian
duke Gebhard, the uncle of the future king Conrad I.

With the premature death of Louis the Child on 24 September 91 1 die east
Frankish line of die Carolinigans came to an end. Only a few weeks later, at all
events before 10 November 91 1, east Frankish magnates set up Conrad the
Younger from the Conradine house as king. He was the first east Frankish king
to be anointed, and was acknowledged widiout difficulty. The Lotharingians
had already defected to die west Frankish ruler Charles the Simple during
Louis’ reign, but the decision to do so was not based on principles of heredi-
tary succession. The main force behind it was the powerful count and missus
Reginar Longneck, who tried in diis way to secure his own claims within
Lotharingia and to exclude his Conradine rivals. It is doubtful if diose east
Frankish magnates who mattered seriously considered the question of
whether to stick widi Carolingian hereditary right and offer Charles the Simple
the succession. Such a decision would have represented more of a break with
tradition than did the election of Conrad, who should be seen as providing
continuity with the east Frankish Carolingians, to whom he was related on his
mother’s side. He dominated the lands around Rhine and Main, the central
region of the east Frankish kingdom, both as duke of Franconia and as the
head of the Conradine family. He had played a significant role in the regency
council and could point to successes in the struggles for supremacy within
Franconia and Lotharingia. His first efforts were devoted to winning back
Lotharingia: these failed, and were abandoned in 9 1 3 . In die same year die mar-
grave Arnulf of Bavaria and the Suabian count palatine Erchanger with his
brodier Berthold fought a victorious campaign against the Magyars; Conrad
was unable to organise the defence against the invaders. It may be drat the new

The Ottomans as kings and emperors

2 39

ethnically based powers already prevented him from doing so; they took up the
task themselves and consolidated dieir position in Conrad I’s reign. In Bavaria
it was Arnulf, and in Suabia it was first of all Erchanger and then the
Hunfriding Burchard II, who became duke of die people. Conrad was deter-
mined to reduce their power. In Suabia he was able to avoid an immediate con-
dict with Erchanger by marrying Erchanger’s sister Kunigunde, the widow of
Luitpold and the mother of Arnulf of Bavaria; once die breach had come in
spite of this, he was able to drive Arnulf out of Bavaria. At the synod of
Hohenaltheim in 916 the pope, in the person of the legate Peter of Orte, inter-
vened decisively in support of Conrad. Whoever rose up against the Lord’s
anointed, so it was decreed, should suffer severe punishments: penances,
excommunication, even execution. Conrad did indeed have the Suabian broth-
ers Erchanger and Berthold executed in 917. Yet die successes trickled away,
the king was defeated, the new duchies established themselves.

Following attacks by both sides a rather different and more promising
arrangement was reached with the Saxon duke, a settlement accompanied by
truce. In 91 5 Conrad probably acknowledged the standing of the Liudolfing
duke and future king Henry as regards his ducal rank, his conquests and what
in effect was his viceregal position. This Franco-Saxon agreement, which even
then may have included a friendship alliance, probably contributed to Conrad’s
proposal that Henry should be his successor, an acknowledgement that a con-
tinuation of Conradine kingship had no future. Conrad, whose power no
longer extended beyond Franconia, died on 23 December 918.

In May 919 Frankish and Saxon magnates elected die Saxon duke Henry in
Fritzlar as their king. Either before or after this, Bavarian and other Frankish
magnates chose the Bavarian duke Arnulf as their king. Duke Burchard of
Suabia and die Suabian magnates did not take part in these elections. These
events offer as it were a snapshot of the east Frankish subkingdoms in action,
as these had been established by die marriage alliances and succession
arrangements made by Louis the German in 872 and continued as indepen-
dent kingdoms after his death in 876. Evidently a king based only on
Franconia as a base domain and on the Rhine-Main area as the central region
of the kingdom was no longer able to sustain a direct and dominant royal lord-
ship extending over several large provinces. In the eastern regna of the former
Carolingian empire, as well as in die west, regional aristocratic forces had
established themselves in positions of leadership with a regal, quasi-regal or
viceregal status. At the end of the Carolingian era it may well have seemed
possible that successor states could be established on this basis; we can appre-
ciate this more clearly if we compare the size of, say, Franconia and Saxony
taken together or of Bavaria alone with the size of the two Burgundian king-
doms or the regnum Italiae.

240

ECKHARD MULLER-MERTENS

THE TRANSLATIO REGNI FRANCORUM AD SAXONES: FOUNDATION AND
CONSOLIDATION OF OTTONI AN FRANCIA ET SAXONIA

The Bavarians set up Duke Arnulf as king ‘in the kingdom of the Teutons’: 2 it
is a matter of dispute whether this view of what happened was already present
in the original Salzburg annals composed around 950 or whether it was intro-
duced at a later time. The now lost original of the annals is preserved only in a
copy produced as an exercise by novice scribes in the middle of the twelfth
century. By then die notion of a ‘kingdom of the Germans’ or a ‘German
kingdom’ had become a commonplace; but when Arnulf and Henry became
kings that was by no means the case. It is probable diat the word Teutonicorum 'vs,
the result of a latter correction or addition, but even if it did occur in the origi-
nal manuscript it still poses the question of what was understood by Teutonici in
Salzburg in the middle of die tenth century: hardly a German people compris-
ing Franks, Suabians, Bavarians, Thuringians and Saxons. There is no evidence
that Arnulf ’s kingship extended beyond Bavaria, and it is probably most easily
understood as a resurrection of the Bavarian kingship practised by Carloman
from 876 to 879. Arnulf was later also to follow in Carloman’s footsteps when
in 934 he intervened in Italy and sought to win the Italian crown for his son
Eberhard.

The Frankish and Saxon electors of the duke of Saxony, Henry, came from
the subkingdom allocated to Louis the Younger in 876. Henry I went beyond
this from the start. Immediately after his election in Fritzlar he took the field
against Burchard of Suabia, who submitted. Henry concluded a friendship-
pact with him, as he had already done with Eberhard of Franconia. The new
Saxon king received Burchard’s submission and at the same time confirmed
the viceregal position of the Suabian duke. Before enforcing acknowledge-
ment in Bavaria Henry turned to Lotharingia, with which he had links through
his sister Oda, the widow of King Zwentibald and the wife of Zwentibald’s
rival, the Matfriding Gerhard. Henry’s accession had coincided with a revolt
against the rule of Charles the Simple in Lotharingia, led by Gislebert, son of
Reginar Longneck, who had died in 9 1 5 . Henry gave Gislebert his support and
intervened in 920 or 921 against Charles the Simple in Lotharingia. The cam-
paigning was ended by an armistice in the summer of 921. Later that year
Henry forced Arnulf of Bavaria to submit. Once more the king made use of a
friendship-pact to define the future nature of the relationship. Arnulf
renounced his royal title and became Henry’s man, but Henry confirmed his
viceregal position. When in 925 Henry finally succeeded in bringing
Lotharingia under his rule he made a similar pact with Gislebert and strength-

2 ‘in regno Teutonicorum’: Annaks ex atmalibus Iuvavensibus antiquis excerpti, s.a. 9 1 9, p. 742.

The Ottomans as kings and emperors 241

ened the relationship by giving Gislebert his daughter Gerberga in marriage in
928.

These friendship-pacts with the dukes were an expression of Henry I’s new
policy, of his intention to redefine kingship. He was more than forty years old
when he became king, and had had extensive experience in establishing his
supremacy and defending his possessions; he was evidently able to recognise
realities and take account of them in pursuing his claims and purposes. The
Liudolfings had established themselves as dukes without meeting any signifi-
cant rivalries, which suggests a degree of consensus with other noble lords,
itself perhaps die product of the somewhat archaic character of the socio-eco-
nomic and socio-political organisation of Saxony. Henry had also experienced
the failure of Conrad’s kingship in the struggle with the duchies. Elected by
Frankish and Saxon magnates as king, the Saxon duke established his lordship
by recognising the intermediate powers in the other provinces and duchies. He
allowed the dukes a viceregal position and bound them to him with pacts of
friendship; at die same time he used vassalitic bonds to subordinate them to
the king. Henry also entered into pacts of friendship with leading noble fami-
lies from Saxony and lower Lotharingia and with die Conradines. The new king
established a relationship of primus inter pares with the magnates, or simply con-
tinued such a relationship from his time as duke. This is probably one of the
reasons why Henry refused to accept the unction which would have set him
above the magnates. But here other factors were at work: Henry intended to
stress a break in continuity with the kingship practised by Conrad I and the east
Frankish Carolingians and to emphasise a new, specifically Liudolfing style of
rulership based on Saxony. The break in continuity was also visible in the fact
that Henry did not take over Conrad’s royal chapel and chancery personnel;
only slowly did he build up a new royal chapel of his own, which from the start
displayed its own, specifically Liudolfing characteristics.

The main feature of Henry’s rule was the recuperation of Lotharingia and
his relations with the rulers of west Francia and Burgundy. He did not seek a
military solution, nor did he allow himself to become involved in factional
infighting, choosing rather to pursue a policy of compromise and reconcilia-
tion while retaining the threat of military intervention as a last resort. In
November 921 he concluded the treaty of Bonn, a pact of friendship with
Charles the Simple, in which the latter appeared as rex Francorum occidentalium
and Henry I as rex Francorum orientalium. Neither before nor after this did
Henry use the east Frankish royal title; his aspirations went further. He was
quite prepared to break the pact with Charles and to enter into another one
with the west Frankish anti-king Robert of Francia at the beginning of 923.
By 925 Lotharingia was already incorporated into Henry’s kingdom, just as it
had belonged to the kingdom of Louis the Younger, whose wife Liudgard was

242

ECKHARD MULLER-MERTENS

an aunt of Henry’s. At an assembly in Worms in 926 Henry concluded a
friendship-pact with Rudolf II of Burgundy. The Burgundian ruler acknowl-
edged Henry’s overlordship, commended himself with the Holy Lance and, in
handing dais over to Henry, also handed over his claims to rule in the kingdom
of Italy. The meeting of die three kings — Henry I, Radulf of west Francia
and Rudolf of Burgundy - at Ivois in 935 demonstrated the predominance
which Henry had reached among die kings of the Carolingian successor-
states; the imperial kinship of the Liudolfing-Ottonian house was now estab-
lished.

The other main feature of Henry’s rule was the defence against die Magyar
invaders and the efforts made to bring die Elbe Slavs under his rule. Already in
the period before his kingship Henry had campaigned against die Daleminzi
on die middle Elbe, who in dieir turn had called on the Magyars for assistance.
It was via the territory of die Daleminzi that the Magyars made their first
attack on Saxony, in 906, and thus defence against the Magyars and control of
the neighbouring areas inhabited by the Elbe Slavs went hand in hand. By
paying a substantial tribute the king was able to purchase a nine years’ truce
from the Magyars; in 927 Duke Arnulf also renewed a truce first concluded in
918. As a result there was an end to Magyar raids from 926 in Saxony and
Franconia, and from 927 in Bavaria, Suabia and Lotharingia. At an assembly
held by Henry with his magnates at Worms in 926, defensive measures were
decided on. These consisted of enlarging existing and setting up new fortifica-
tions, and were to apply diroughout the kingdom. In Saxony, whose military
obligations still reflected those of an archaic period, free peasant warriors
( agrarii milites) were entrusted both with the building of die fortifications and
widi their subsequent garrisoning. Henry also took steps to increase the
number of mailed horsemen available, in other words to modernise the Saxon
army. The new troops were tried out in the campaigns of 928-34 against the
Elbe Slavs, where they played a significant role. The territories of the Slav
peoples as far as the Oder were largely subjected to a still impermanent over-
lordship and to tributary dependence. In 929, furthermore, Henry compelled
Duke Wenceslas of Bohemia to submit. In the territory of the Daleminzi he
established a centre for his lordship in the fortification at Meissen. As com-
manders we find Count Siegfried in Merseburg and the margraves Bernard and
Thietmar on the lower Elbe: in 929 these defeated a Slav army which had
advanced across the Elbe at Lenzen, a batde in which mailed horsemen played
a crucial role. By 932 Henry felt strong enough to risk open conflict widi the
Magyars; he cancelled the tribute payments. When in 933 the Magyars
appeared in Saxony in response to this affront, drey were defeated by an army
whose main component was heavily armed cavalry and which consisted of
levies drawn from all the gentes of east Francia. In the following year Henry

The Ottomans as kings and emperors 243

defeated a Danish petty king on the northern frontier. The victories over both
invaders, the Magyars and the Norsemen, brought the first Saxon king a repu-
tation which extended well beyond the borders of his kingdom.

Henry’s practice of kingship reflected both his own aspirations and the
changes which had taken place in the structure of the kingdom. The Saxon
king did not set foot in Bavaria and Suabia after the submission of their
dukes; 3 these remained distant zones in relation to his kingship. The investi-
ture of the Conradine Hermann as duke of Suabia in 926 made no difference
to this. The central political region of the former east Frankish kingdom, the
area around Rhine and Main, was now joined by east Saxony and Thuringia as
a new base domain. In addition the area between the lower Rhine and Meuse
centred on Aachen took on a new role as a region for the exercise of kingship.
This area, a core domain in the Carolingian empire, had never been of central
importance for the Carolingian east Frankish kingdom. When we consider its
structure in this way, Henry I’s kingdom appears not so much a continuation
of Francia orientals (the east Frankish kingdom with its core regions around
the Rhine— Main confluence and in Bavaria linked by a transit region consist-
ing of Suabia and east Franconia) as a renewal of the political constellation
established by Charles the Great, in which the core regions had been those
around lower Rhine and Meuse, around Rhine and Main, and in Saxony,
linked by transit regions in Westfalia and in Hesse and east Franconia. It was
Henry I’s practice and the political reorganisation during his reign which
established the Ottonian kingdom of Francia et Saxonia, acknowledged in
diplomata issued by Otto I in 936 and 938 as die provinces represented by his
kingship. 4

The first Saxon king saw himself as a rex Francornm, but he did not follow the
dynastic practice of the Franks by dividing the kingdom between his heirs. The
kingship which the Liudolfings had won was to remain indivisible. At the
assemblies of 929 in Quedlinburg and 936 in Erfurt Henry’s first son by his
Immeding wife Matilda was chosen as his successor. The older son by his first
marriage, Thankmar, and the younger son by Matilda, Henry, who had been
born after Henry had already become king, were given no part of the kingship.
Henry here followed a general trend becoming visible both in the practice and
in die new conceptions of the state which were developed in the Carolingian
successor-states and in die aristocratic principalities. For the kingdoms as
feudal states, as the form of political organisation taken on by an aristocratic
society, it came to be established that their basis was the community of aristo-
cratic and ecclesiastical magnates. The kingdom was conceived of as existing

3 The visits to these provinces posited by Schmid (1964), pp. 113—22 for 929/ 30 have no direct basis in

the sources; the Liudolfing entries in the Suabian libri memoriales do not require us to assume that

Henry I was present in person See Althoff (i 992), p. 1 1 1 . 4 DD O 1 1 and 20.

244

ECKHARD MULLER-MERTENS

apart from the ruling family, and the royal office was contrasted widr the
person of the king. To preserve the unity of die kingdoms evidendy corre-
sponded to the interests and intentions of the new non-Carolingian dynasties
as well as of die princes and die major churches. On 2 July 936 King Henry
died in die palace at Memleben. Later generations saw him, according to
Widukind, as the greatest of the kings of Europe, who left his son a great and
broad kingdom which he had not inherited from his fathers but acquired by his
own efforts and by God’s grace. 5

Only five weeks later, on 7 August 936, Otto I was raised to kingship in
Aachen by all the dukes and other magnates, and was crowned and anointed by
the archbishops of Mainz and Cologne. Otto appeared in Frankish clothing
and was enthroned both on the throne in the arcaded court in front of the
Aachen palace chapel and on the dirone in the chapel’s upper storey. Henry’s
successor had thus taken possession of Charles the Great’s throne in a manner
so visible as to leave no doubt. It was a politically motivated and programmatic
act, one which should be seen in a line of the continuity with Henry I’s king-
ship. Liudolfing (henceforth Ottoman) kingship did not merely stand in suc-
cession to the east Frankish kingdom of Louis the German. It had now
established itself in succession to Charles the Great in Aachen, in a core
domain of great importance for the self-definition of Charles the Great’s
empire, die area around the lower Rhine and Meuse which had been associated
widr the imperial title in the division of 843 .

The new king proceeded without delay from the Carolingian centre of
Aachen to the Ottoman centre at Quedlinburg, Henry I’s principal palace and
his burial place. The nunnery founded there by Queen Matilda, Otto’s mother,
received a rich endowment. Otto moved against the rebellious Elbe Slavs as
early as September 936. New margravates were established on the lower Elbe
and on the middle Elbe and Saale, a sign of a policy of more intensive domina-
tion and of an intention to claim and establish hegemonial kingship over the
west Slav areas up to the Oder. The foundation of the monastery dedicated to
St Maurice at Magdeburg in September 937 was a further sign of this. The new
royal foundations in Quedlinburg and Magdeburg were significant for Otto’s
programmatic conception of kingship in another respect as well. Quedlinburg
was dedicated not only to Peter and Mary but also to the Maastricht saint,
Servatius, and to the west Frankish royal saint, Denis. The new monastery at
Magdeburg was staffed by monks from St Maximin’s in Trier and had St
Maurice and his companions as patrons. Relics of one of die companions, St
Innocentius, had been donated by the king of Burgundy, Rudolf II, who had
just died and whose minor son and successor Conrad had been taken under

5 Widukind, Res gestae Saxonicae 1,41.

The Ottomans as kings and emperors 24 5

Otto’s control. These actions with their associated relics and dedications estab-
lished links with Lotharingia, Burgundy and west Francia; they implied an
imperial kingship. Otto himself described his kingship in the diploma for
Quedlinburg (D 0 1 1) as being based on the royal throne in Francia ac Saxonia, a
throne which he saw as something separate from his own dynasty and at the
disposal of the electors.

The appointment of Hermann Billung as margrave on the lower Elbe and of
Gero on the middle Elbe and Saale infringed die claims of Hermann’s brodier
Wichmann I and of Otto’s half-brother Thankmar. Otto was equally unpre-
pared to accept Eberhard of Franconia’s behaviour towards a Saxon vassal,
and later on he rejected an agreement which Archbishop Frederick of Mainz
had made on his behalf. He thus injured duke and archbishop in their dignity
and reputation. On the death of Duke Arnulf of Bavaria he reduced the extent
of the ducal rights enjoyed by his sons. These incidents produced waves of dis-
approval, of outrage; they affected the balance of power between and the
ranking of the aristocratic kindreds, connections and communities, and those
offended were prepared to defend their claims with feud. An early failure by the
young king against the sons of the Bavarian duke in 93 8 triggered off a series of
uprisings in Saxony, Franconia and Lotharingia which lasted until 941. The
heads of the rebellions and conspiracies were members of the royal family:
Thankmar, who was killed in 938, and then Otto’s younger brother Henry.
Duke Gislebert of Lotharingia sought his own advantage in supporting Henry.
Henry wanted to dethrone Otto and become sole ruler himself. The Suabian
duke Hermann and other members of the Conradine family supported Otto.
The issue was decided in favour of the king at Andernach on the Rhine in
September 939. The resistance and revolt offered by substantial parts of the
church and the lay aristocracy were rooted in the claim by die new ruler to an
enhanced kingship. However, it should not be overlooked that at that time
every change of ruler produced tensions, and disruptions of die pecking order
and of the possessions and intiuence of the lay and ecclesiastical magnates.
Even if die succession question had been decided otherwise, even if die rela-
tionship with Eberhard of Franconia had taken a different course in 936 and
937, the new ruler of an undivided kingdom would still have been faced with
condicts and trials of strengdi. Widiout a predominant central power, which in
turn required a sacral basis and legitimation, a transethnic, hegemonial and
imperial kingship of the kind established by Henry I and passed on undivided
to his son could neither be maintained nor made permanent. Otto broke with
his father’s practice of presenting himself as first among equals. He was not
prepared to enter into amicitiae, into agreements with other magnates which
placed obligations on both sides. He insisted on submission and made deci-
sions arbitrarily after the fashion of Carolingian kingship. He demanded the

246 ECKHARD MULLER-MERTENS

precedence due to him as the Lord’s anointed; he was convinced of the nature
of his kingship as kingship by the grace of God.

Otto did not appoint a new duke in Franconia after Duke Eberhard had
been killed at Andernach together with Gislebert of Lotharingia. Like the
Saxon ethnic dukedom, die dukedom of the royal province of Franconia was
taken under direct Ottoman rule. Otherwise Otto pursued his father’s policy of
recognising the dukedoms established over edinic areas and royal provinces as
established parts of the kingdom’s organisation. As a counterpart to this he
continued with the practice, already found under Arnulf of Carinthia and reac-
tivated by Henry I, of entering into relationships of feudal overlordship with
non-Carolingian rulers in the Carolingian successor-states. Otto introduced a
new element into this policy. He sought to link the royal dynasty with those of
the dukes by marriage alliances and to secure the dukedoms for members of
the royal family. In Bavaria Duke Berthold, Arnulf ’s brother, was intended to
be married to Otto’s sister Gerberga or her daughter. The marriage did not take
place, but Otto’s brother Henry succeeded Berthold, and he had been married
to Arnulf ’s daughter Judith since 936 or 937. In Suabia Otto set up his son and
heir-presumptive Liudolf as duke; he had been married to Ida, the daughter of
Duke Hermann, who had no sons, in 947. In the same year Otto’s daughter
Liudgard was married to Duke Conrad the Red of Lotharingia. Following
Conrad’s deposition in 95 3 Otto made his brother Brun, who at the same time
became archbishop of Cologne, duke of Lotharingia.

From the beginning of his rule Otto was locked in to west Frankish,
Burgundian and Italian politics as a consequence of existing relationships. This
involved disputes over kingship, transregnal aristocratic connections and con-
flicts between kings and aristocratic factions. Otto, who disposed of superior
forces and armaments, soon found himself in a hegemonial position. In west
Francia he encouraged a balance between King Louis IV and Duke Hugh of
Francia, both of whom were married (from 939 and 937 respectively) to sisters
of Otto: Gerberga, Gislebert’s widow, and Hadwig. Otto met Louis no fewer
than seven times between 942 and 950, for example at the general synod of
Ingelheim in 948, which settled the schism in the archbishopric of Rheims and
allowed Otto to display himself in the full glory of his hegemonial kingship in
alliance with tire papacy. After the deaths of Louis and Hugh their widows,
Otto’s sisters, managed Carolingian and Robertine power in west Francia, and
their brother Brun, archbishop of Cologne, was de facto regent. Conrad of
Burgundy had grown up at Otto’s court. In 942 he reached the age of majority,
and with Otto’s backing was able to achieve a kingship extending over
Provence as well as Burgundy. As to Italian affairs, Otto provided from 941
onwards a refuge for Margrave Berengar of Ivrea, the rival for the throne to
die kings, Hugh and Lothar. Berengar returned to Italy in 945 with Otto’s

The Ottomans as kings and emperors 247

agreement, drove out Hugh, who died in 948, and was a serious direat to the
kingship of Hugh’s son Lothar. Lothar’s death in 950 established a new politi-
cal situation. Berengar had himself crowned king of Italy, imprisoned Lothar’s
widow, Adelaide of Burgundy, and so provoked independent interventions in
Italy by Otto’s brother Henry of Bavaria and his son Liudolf of Suabia. The
disturbances this created in the balance of power in Italy, Burgundy and south-
ern Germany compelled Otto to intervene in Italy himself.

During his first Italian expedition, between the autumn of 951 and February
952, Otto received kingship over Italy and married Adelaide, the dowager
queen. But he was able neither to drive out Berengar II nor to organise an expe-
dition to Rome to receive imperial coronation. The collisions of interest and
decisions in and around Italy provoked new disputes within Otto’s own family,
which developed into a new general uprising against Otto’s kingship. Conrad
the Red, who had remained in Italy, made an agreement with Berengar that the
latter should retain die kingship over Italy under Otto’s overlordship. Otto was
furious, and wanted to institute proceedings for high treason; but the agree-
ment held. But before Berengar was finally able to receive investiture with the
kingship in return for homage at the assembly in Augsburg in 952, Otto had
ostentatiously demonstrated his displeasure with Conrad and Berengar at the
Easter celebrations that year. For Conrad diis was a reason to join Liudolf ’s
conspiracy; Archbishop Frederick of Mainz had already done so.

Liudolf, who had himself had hopes of the crown of the Lombard
kingdom, found not only that he was excluded in Italy while his uncle Henry
was favoured, but also that his own succession was threatened. He set himself
up in opposition to the excessive influence of Henry and the new queen at
court. These conflicts revived the latent tensions between Otto’s claims to
kingly rule and die claims of lay and ecclesiastical magnates to participate in
government. When in 95 3 Otto rejected an agreement which had been nego-
tiated at Mainz between himself, Liudolf and Conrad by Frederick of Mainz
on the grounds that it had been extorted, Liudolf ’s rebellion broke out openly;
it lasted until 955. Liudolf, Conrad and Frederick could rely on existing aristo-
cratic friendship agreements which implied obligations of mutual assistance.
Liudolf was backed by Suabia; the old conflicts with the church of Mainz and
with the Saxon nobility broke out afresh; opposition within Bavaria to Henry’s
rule became visible. Otto was not able to suppress die rebellion by military
force. The Magyars renewed their attacks on the Reich in 954, and the
Abodrites east of the Elbe rebelled. The revolt collapsed in the face of the
threat from the invading Magyars, especially after Conrad and Liudolf had
seemed to ally themselves with the invaders. The rebels submitted; Conrad and
Liudolf lost their duchies, and the last centre of resistance, Regensburg, sur-
rendered in April 955.

248 ECKHARD MULLER-MERTENS

When in 95 5 die Magyars again raided into Bavaria and Suabia, Otto met
them with an army of Franks, Suabians, Bavarians and Bohemians. The
Lotharingians were absent because of die long distance, and the Saxons were
pinned down by condicts with the Elbe Slavs, who had been joined in the
course of Liudolf’s uprising by the Billungs, Wichmann the Younger and
Ekbert. Otto, who led his army into combat carrying the Holy Lance, forced
batde on the Hungarians while drey were besieging Augsburg. They were
crushingly defeated on 1 o August 9 5 5 at the Lechfeld and on die two following
days while deeing. Immediately after his triumphal victory against the
Hungarians Otto moved against the Abodrites, who were defeated at the
Recknitz in eastern Mecklenburg on 1 6 October 955.

Before the batde the king received an offer of terms from the Abodrites:
they were willing to pay tribute, but they wanted to retain freedom and lordship
over their own land. 6 Otto’s conception was one of direct rule over the neigh-
bouring Polabian and Sorbian setdement areas. He aimed at immediate lord-
ship over land and men, accompanied by renders and services from the rural
population and ultimately the formation of manorially organised estates. To
this end a military and political organisation was set up, in die first instance in
the form of the wide-ranging marches controlled by Hermann Billung and
Gero. On Gero’s death in 965 the march on the middle Elbe and on the Saale
was divided into six smaller marches. Besides the marches there were smaller-
scale organisational units in the lands of the Sorbs and Hevelli; in the Latin
diplomata of the time these are denoted by their Saxon name of burgward. A
burgward comprised some five to twenty villages grouped around a fortification
which was a centre of administration and lordship. The fortifications were to
be erected and maintained by the Slav population. Alongside these organisa-
tional units of secular lordship went a new church organisation. The bishop-
rics of Brandenburg and Havelberg were set up in 948, and were joined from
967 onwards by bishoprics at Meissen, Merseburg and Zeitz and at Oldenburg
in east Holstein. Otto planned from an early date to transform the monastery
of St Maurice at Magdeburg into a new archbishopric which should be the
centre of church organisation and missionary activity for the conquered Slavic
territories. He sought the pope’s agreement to the foundation of bishoprics at
Magdeburg and elsewhere immediately after die batde on die Lechfeld, but
this idea was wrecked by the opposition which came from the church of Mainz
under its new archbishop, William, a son of Otto’s by a Slav concubine of royal
descent.

The structures established by Henry I were consolidated by Otto I and con-
tinued under his successors. Compared with the east Frankish Carolingian

6 Widukind, Res gestae Saxonicae, in, 53 .

The Ottomans as kings and emperors

249

kingdom, die Ottoman Reich had a new central region and base domain in the
lands around the Harz in eastern Saxony and northern Thuringia. A second
central region lay in the lands around die confluence of Rhine and Main; this
had been established in the time of Charles the Great and had been the princi-
pal central region of the east Frankish kingdom. Henry I and Otto I reactivated
a third central region, that established by Charles the Great in the lands around
lower Rhine and Meuse with its centre at Aachen, as a core domain of the
Ottoman Reich. In all three central regions Otto exercised direct royal rule at
periodic intervals, and this was accompanied by regular transit passages and
assemblies in Angria/Westfalia and in Hessen/Thuringia/ east Franconia. It
was in die central regions above all that the assemblies took place and die prin-
cipal acts of rulership, the religious representation of die ruler on the high
feasts of the church, implying a ritualisation of the ruler by die grace of God
which displayed his sacrality and his God-given nature. The remaining political
regions - Bavaria, Suabia, Alsace, upper Lotharingia, Frisia - were visited only
occasionally by Otto, either in the course of rebellions or in transit to Italy or
west Francia. The dukes, counts, bishops, abbots and other magnates of these
distant zones met the king either in one of die central regions or in the east
Franconian transit region. As a result each region developed its own catchment
area for the lay and ecclesiastical magnates of the kingdom, and this led to the
creation of supra-ethnic and supra-regional sets of connections based on the
central regions, each with its own infrastructure of roads and supplies. The
main east-west lines of supply and communication were those established by
Charles the Great, either the Hellweg from the lower Rhine to Saxony or those
roads running from the middle Rhine to Saxony via Main, Hesse and east
Franconia. To this the Ottomans added a north-south axis which linked east
Saxony/north Thuringia via east Franconia with southern Germany. The exis-
tence of such structures does not permit us to call the Ottoman Reich a contin-
uation of the regnum Francorum, even though contemporaries like Widukind of
Corvey used this term for it and conceived of it in this way. The Ottoman Reich
displayed new Saxon characteristics in its political structure over and above die
ones which it had inherited.

The phrase Francia et Saxonia, used by Otto I in 936 and 938 and later by
Widukind of Corvey and Adalbert of Magdeburg, corresponds better to the
actual structure of the Reich. The reality of Ottoman kingship in the period
before Otto’s imperial coronation was reflected in the works of writers of the
960s, who talked of the populus Francorum et Saxon urn and developed the idea of
a ‘translation’ of the Frankish empire to the Saxons. But there was also a spe-
cifically Saxon conception of the Reich, which could be seen as a regnum
Saxonum.

250

ECKHARD MULLER-MERTENS

OTTO THE GREAT AS EMPEROR AND THE APOGEE OF THE OTTONIAN

REICH

It is only following the imperial coronation of Otto I in 962 that an Ottoman
historiography can be said to begin; from the preceding period we have only
Liudprand of Cremona’s Antapodosis. There had been no serious historio-
graphical activity since around 900, but now as well as Liudprand we have
Widukind of Corvey, Hrotsvitha of Gandersheim, Adalbert of Weissenberg,
Ruotger and the author of the older \ 'ita Mathildis as well as Rather of Verona
and a number of significant hagiographers. Most of these authors saw Otto I’s
imperial coronation as marking the beginning of an epoch and the papal coro-
nation as constitutive for the imperial quality and dignity. The imperial aspect
of his kingship was stressed for the period before the coronation, something
which had already been done by the royal chancery during Otto’s kingship.
Some of the historians even projected the imperial tide back into the royal era.
Apart from Liudprand and Rather, only Saxon authors of this period gave the
Ottoman emperorship a Roman name. This took place before Otto II actually
assumed the Roman imperial tide in 982, and suggests a stress on the Roman
element of emperorship which was not merely literary or historiographical.
Widukind by contrast stressed a non-Roman imperial idea: he derived Otto’s
imperial rank from an acclamation as imperator by his army following the
victory on the Lechfeld in 9 5 5 . The idea of a non-Roman emperorship is also
implicit in an ordo for die coronation of an emperor composed c. 960 in Mainz,
probably in the entourage of Archbishop William. Like Widukind, William was
an opponent of Otto’s project to establish an archbishopric at Magdeburg with
the help of the pope and of his imperial tide.

From the beginning Otto had practised a hegemonial and imperial kingship.
It expressed itself in supremacy and hegemony over the other Carolingian suc-
cessor-states and in a policy of political and military expansion against the
Slavonic lands to the east; here and in the north it was also linked with mission
and the development of church organisation. Otto had already negotiated
about an imperial coronation in 951. The Roman patricius Alberic, who had
Rome and die patrimonium Petri firmly under control, opposed the request.
Following Alberic’s deadi in 9 5 4 the political constellation in Rome shifted. In
955 Agapetus II consented to the foundation of an archbishopric at
Magdeburg, but the plan failed against the opposition by William of Mainz. On
this occasion too Otto may have put out feelers about an imperial coronation.
It was only very reluctandy in 9 5 2 that he had recognised Berengar II as Italian
king under his overlordship. When he was once again master of the situation
after the suppression of Liudolf ’s revolt, he sent Liudolf into Italy to drive out
Berengar II and his son Adalbert, who had become effectively independent;

The Ottomans as kings and emperors

251

Liudolf had the prospect of himself becoming subking in Italy. Following
Liudolf’s death in the autumn of 957 the trwo recovered and extended their
power. They came to threaten the pope, who was also menaced by die princes
of Benevento and Capua and by opposition within Rome itself. John XII
invited King Otto to free the Roman church and the pope from die tyranny of
Berengar and Adalbert. The new disturbance of the balance of power within
Italy and the inner-Italian threat to the papacy gave Otto the opportunity to
acquire the emperorship. Just as his Carolingian predecessors Pippin and
Charles had responded to die calls for help by the popes, Stephen II, Hadrian I
and Leo III, so Otto responded to John XII’s appeal. He made arrangements
for the period of his absence, ordered the succession and had his son Otto II
elected and crowned as co-king. On 2 February 962 Otto received the imperial
crown from Pope John XII, and Adelaide was crowned and anointed widi him.
After John XII had agreed to die setting up of an archbishopric in Magdeburg
and given Otto a free hand to organise die church in the Slavonic east, die new
Ottoman emperor confirmed Pippin’s donation of 7 5 4 and the other privileges
granted by the Carolingians to the pope and to the Roman church, and secured
for himself the imperial right to a promise of fidelity from the pope after he
had been elected by the nobility and clergy of Rome and before he was conse-
crated. Otto assumed neither Charlemagne’s imperial title with its reference to
the Roman empire nor an imperial title which referred to the Romans; he was
content with the simple tide inrperator augustus customary from the time of
Louis the Pious onwards. In die chancery the designation magnus for Otto
became customary after 962; he has gone down in history with this honorific
title, ‘the Great’.

Following the imperial coronation Otto began warfare against the kings,
Berengar and Adalbert, and continued this with interruptions and pauses.
Berengar surrendered at the end of 963 and was sent into exile at Bamberg.
Adalbert fled, returned after the emperor’s departure from Italy, and was
defeated by an army sent by Otto I in 965 under Duke Burchard of Suabia; only
after diis did he cease to be a factor in Italian politics. Otto did not appoint a
new subking, and ruled directly in the Italian kingdom from the autumn of
961. Otto and his successors employed a policy of exploiting aristocratic con-
flicts of interest, of preserving the balance of power between margraves,
counts and bishops, of encouraging die development of new smaller margra-
vates and granting privileges to the bishoprics. In this way they were able to
stabilise dieir rule in the regnum Italiae. The main problem following die impe-
rial coronation was to setde the relationship between emperor, pope and
Rome. The Roman aristocracy and their aristocratic pope had no interest in a
serious practical application of imperial power over Rome and within Italy.
In the period after the coronation there was a series of uprisings and judicial

252

ECKHARD MULLER-MERTENS

hearings with rapid changes of control over Rome. Typically for his style of
rulership, Otto demanded submission to his will in an intensified and previ-
ously unknown claim to rulership. After John XII’s rapid defection Otto had
him deposed by a Roman synod and an antipope, Leo VIII, elected, here
infringing the legal principle drat a pope is not subject to any earthly tribunal.
The Romans also had to take an oath to Otto that they would never elect and
consecrate a pope without permission from Otto and his son. John won Rome
back, and a new Roman synod condemned Leo’s election. Following John’s
death in May 964 the Romans asked Otto for his consent to the election of a
new pope, as they did not recognise Leo VIII. The emperor insisted on having
his pope. He besieged Rome in June 964 until the Romans had handed over
their pope, Benedict V, who was kept prisoner in Hamburg until his death.
Otto had thus compelled papacy and Romans to acknowledge a genuine exer-
cise of imperial power. At the beginning of 965 he returned to the German
lands of his empire.

It was not until 967/ 8, during the third Italian expedition, that the emperor
managed to have an archbishopric set up at Magdeburg. The new province of
Magdeburg included the existing bishoprics of Brandenburg and Havelberg as
well as the newly erected sees at Merseburg, Zeitz and Meissen. The new foun-
dation at Oldenburg, like the bishoprics of Schleswig, Ribe and Aarhus estab-
lished in the course of the Danish mission which Otto had patronised, came
under the archbishopric of Hamburg-Bremen. The resistance to tire establish-
ing of an archbishopric at Magdeburg, led by William of Mainz and Bernard of
Halberstadt, shows the practical limitation of royal power by institutions like
the churches of Mainz and Halberstadt and also reveals communities of inter-
est between these and groups of magnates. Only after die death of Bernard in
February and of William in March 968 was Otto finally able to complete the
foundation of Magdeburg in cooperation with Pope John XIII at a synod at
Ravenna in October 968. It is still disputed whether the bishopric of Posen,
also founded around this time, belonged to Magdeburg’s province or not. Its
existence hints at the rise of a new concentration of power to die east in the
form of the nascent Polish state and its church organisation. Margrave Gero
had concluded a pact of amicitia on Otto’s behalf with the increasingly power-
ful Polish prince Miesco in 963, which included the obligation to pay tribute for
the western part of the Polish realm as far as the Wartha. In 966 Miesco
received baptism, and the bishopric of Posen was founded shortly afterwards
in the heart of his realm: a new Christian power was establishing itself firmly.

Otto’s brother Brun, who had simultaneously been archbishop of Cologne
and duke of Lotharingia, died in 965 . He is held to have been a prototype of the
Ottoman imperial bishop, the inaugurator of an Ottoman imperial church
system. The imperial church itself was a Carolingian inheritance. By contrast

The Ottomans as kings and emperors

253

with the situation in the west Frankish kingdom the east Frankish rulers had
retained lordship over all the bishoprics and over many monasteries, so that
there was here a much greater element of continuity between the Carolingian
and tire Ottoman eras. This imparted a special flavour to die church in the
Ottoman Reich and in the end led to the development of a specifically
Ottoman and Salian imperial church; Brun’s activities lie at the beginnings of
this. He combined and conflated episcopal and ducal office in the service of the
Reich and had clerics drawn from the high nobility for the service of church
and Reich trained at the cathedral school of Cologne, whom he then had
appointed to Lotharingian bishoprics. Otto took over this conception, or
rather it fitted in with his existing style of rulership. He extended the royal
chapel, and staffed it above all with cathedral canons drawn from the high
nobility; he increased the proportion of chaplains promoted to bishoprics.
This began before Brun’s death, and indeed tire beginnings of tire practice can
be traced to die reign of Henry I, who had established a new royal chapel and
appointed cathedral canons as capellani. Brun himself had been educated at the
Utrecht cathedral school, and served from 941 to 95 3 as chaplain and chancel-
lor before he became archbishop. What was to become die characteristic
feature of die imperial church in the Ottoman and Salian eras - the intercon-
nections between royal court and royal chapel and the cathedral chapters, and
bishoprics held by former chaplains - set in before 967 and was not linked with
specific events like Liudolf ’s uprising or the death of Brun. A second charac-
teristic was the religious legitimation of the widespread and traditional practice
that tire bishops performed secular services for the king. These stood in contra-
diction to the episcopal ideal established in the course of the ninth century,
which was orientated towards Benedictine monasticism and a monastic-ascetic
way of life. In his vita of Brun, Ruotger justified the involvement of bishops
‘with political affairs and dangerous wars’. 7 He worked out a new episcopal
ideal, which Brun incorporated. Service for the king, who for contemporaries
was the image of God and Christ’s representative as well as tire defender of the
church, hence a priest-king, was here depicted as a duty which conformed with
the divinely willed order of tilings. Service for the Reich — a part of the divine
order of tilings, entrusted to the ruler by God - was die service of God.

A new Roman uprising against Pope John XIII recalled Otto to Italy, this
time for what was to be a stay lasting six years, from 966 to 972. Otto, drawn by
his Roman policy into the relations between the papacy and the Lombard prin-
cipalities, came into conflict with Byzantium about the disputed overlordship
over the latter. Finally war broke out, with a number of unproductive expedi-
tions by Otto into Byzantine southern Italy. The result was a compromise:

7 ‘rem populi et pericula belli’: Ruotger, 1 Vita sancti Brunonis archiepiscopi Coloniensis , c. 2 3 .

254

ECKHARD MULLER-MERTENS

Otto retained feudal overlordship over Benevento and Capua, and his western
emperorship was recognised by the east Roman empire. He also secured a
Byzantine princess - Theophanu, the niece of the new, peaceably inclined
emperor John Tzimisces — for his son, who had been crowned co-emperor in
967. The brilliant marriage of Otto II and Theophanu in St Peter’s in Rome in
972 displayed Otto I at die height, diough also at the limits of his power; there
were rumbles of discontent in Saxony at this time over the ruler’s long absence.
The emperor returned; he died on 7 May 973, like his father in the palace at
Memleben.

OTTO II AND OTTO III: MILITARY DEFEATS AND THE COLLAPSE OF
THE POLICY OF RENOVATIO IMPERII ROUANORUU

Otto II had already been chosen king and emperor during his father’s lifetime.
He succeeded Otto I, aged not quite eighteen years, without a new election as
king or a new imperial coronation in Rome. The questions of pecking-order
and claims to power of the magnates associated with die succession soon led
to conflicts and trials of strength in the south German duchies and in
Lotharingia. In the west diese centred around the claims of those magnates
who had been driven into exile in 958 and now returned. Otto II restored
Reginar IV and Lambert, the sons of Reginar III, to their allodial lands and in
977 appointed the Carolingian Charles, who was at loggerheads with his
brodier, die west Frankish king Lothar, as duke in lower Lotharingia. The con-
sequence was war with Lothar, which was ended in 980 by an agreement to
restore the status quo. In Bavaria Henry die Quarrelsome, die nephew of Otto
I and cousin of Otto II, togedier widi the Luitpoldings, revealed their ambi-
tions; Otto responded by favouring his other nephew Otto, the son of Liudolf,
and granting him Suabia in 973 and Bavaria in 976. Henry II of Bavaria tried to
depose his cousin, and organised several uprisings in alliance widi Miesco of
Poland and Boleslav II of Bohemia. In the course of the struggle Carindiia was
separated off from Bavaria and made into a duchy in 976, the Main-Frankish
family of the Babenberger were given the Bavarian nordiern march, set up
after 955, and finally in 978 Henry II lost his duchy. In these years the most
induential advisers at the royal court were the empress Theophanu, Willigis,
the last chancellor of Otto the Great and from 975 archbishop of Mainz, and
Hildibald, the chancellor and bishop of Worms.

Royal or imperial rule remained undisturbed in Italy after Otto I’s deadi, but
not in Rome. Roman aristocratic factions, among whom die Crescentii were
the most powerful during the second half of the tenth century, and popes
established eidier by the emperor or by die Crescentii continued fluctuating
struggles for power. At the end of 980 Otto II left for Italy. Theophanu, hostile

The Ottomans as kings and emperors 2 5 5

to the new regime in Byzantium following the change of dynasty in 97 6,
together with Gerbert of Aurillac and Adso of Montier-en-Der, introduced
the young emperor to the Roman imperial conception and the idea of bringing
the whole Italian peninsula under Roman imperial lordship. In Rome he
decided to drive back the Saracen offensive in the south Italian mainland and in
this way to conquer the Byzantine areas of southern Italy. In order to
strengthen his army Otto ordered 2100 additional mailed horsemen from
Germany, of which around 1 500 were to be provided by the imperial churches.
In the course of the siege of Tarento in March 982 die emperor adopted the
Roman imperial title and thus proclaimed his claim to rule over a Roman empire
against that of Byzantium. The campaign in southern Italy ended in a disaster:
at Cotrone on the Calabrian coast the imperial army was crushingly defeated by
the Saracens injuly 982.

Saxon magnates demanded a meeting with die emperor, who summoned an
assembly to Verona in May 983. Since Duke Otto of Suabia and Bavaria had
died in Italy, the south German duchies were vacant. The vacancies were filled
by members of the old ducal families: Bavaria reverted to the Luitpoldings in
the person of Henry the Younger (who had been deposed as duke of Carinthia
in 978), while Suabia was granted to tire Conradines in the person of Conrad, a
nephew of Duke Hermann. A further crucial matter dealt with at the assembly
was tire settling of the succession. The emperor’s three-year-old son, Otto III,
was elected as king with the participation of die Italian magnates present. The
young Otto was sent to Germany to be crowned at Aachen by the archbishops
Willigis of Mainz and John of Ravenna and to be brought up by Warin, arch-
bishop of Cologne. Otto’s election in Verona by German and Italian princes,
and his coronation by the archbishops of Ravenna and Mainz, show the wish
of the court and the participating magnates to treat the German territories and
the regnum Italiae as one kingdom, to stress tire one imperial kingship which
integrated the various regna. Otto II had made intermittent use of the Roman
imperial title in the course of the struggle with Byzantium. Following die
Veronese assembly he renewed his attempt on southern Italy; in the course of
this he died at Rome, on 7 December 983.

The news of the great uprising by the Elbe Slavs probably reached Otto
before his death. The Liutizic confederation formed in the preceding years, an
alliance of the Elbe Slav tribes with the Redarii at its core, had risen in the
summer of 983 together with the Abodrites against Ottoman lordship. The
Saxon march and church organisation was swept away. The Abodrites burned
Hamburg. The attack was checked only west of the Elbe at the Tanger, where a
Saxon army defeated the insurgents. The defeat at Cotrone and the Elbe Slav
uprising brought the Ottomans and the Reich they had founded their first
serious defeats. The extension of Ottoman power into southern Italy, already

256 ECKHARD MULLER-MERTENS

checked under Otto I, was halted once again, and Saxon expansion into the
areas of Polabic tribal organisation and setdement was reduced to its starting-
point at the Elbe.

The dispute about the guardianship over the royal child Otto III (claims
were made by Henry the Quarrelsome and tire west Frankish king Lothar) or
indeed about the succession (Henry the Quarrelsome had himself elected king
at Easter 984) ended in favour of Empress Theophanu. She was supported
above all by Archbishop Willigis of Mainz, who, together with Hildibald of
Worms, continued to exercise decisive influence at court. This did not change
when die regency was taken over by Empress Adelaide following Theophanu’s
death on 15 June 991 at Nijmegen. The basis of royal rule remained intact
during the years of regency under the two empresses, both in the German
lands and in the kingdom of Italy; Adelaide exercised royal authority in Italy for
years. Henry tire Quarrelsome was restored to Bavaria in 985, and later to
Carinthia. The structures and style of rulership, as visible in tire itineraries of
the empresses and the grants of privileges, corresponded to those practised in
the time of Otto the Great. Promotions of royal chaplains to bishoprics con-
tinued, and in 985 Theophanu made the first grant of a whole county to a bish-
opric, Liege.

In 989 and 990 Theophanu made a journey to Italy. She exercised imperial
power in Rome and Ravenna, and issued diplomata as Theophanu inrperatrix
augusta , even as Theophanius wrperator augustus, but avoided getting involved in the
Roman factional disputes. In spite of her involvement in west Frankish affairs
and in tire rivalry between Capetians and Carolingians for the throne, which cul-
minated in the election of Hugh Capet in 987, her activities did not go beyond
securing Lotharingia with diplomatic means. The renewed claim of the west
Frankish Carolingians to this were rejected, and Hugh Capet, supported by
Theophanu in his bid for tire throne, as newly-elected king of France renounced
Lotharingia. A comparison of the Rheims electoral dispute of 989-97 with that
of 940-8 clearly shows how France and the Ottoman Reich had developed away
from each other. In the latter dispute the influence both of the papacy and of
the Ottoman court were less visible; it was Hugh Capet who dominated. The old
imperial position of tire Ottomans no longer had a basis in France.

In almost every year of her regency Theophanu organised campaigns in the
Elbe Slav territories and often participated herself. These were conducted in
alliance with Duke Miesco of Poland, who did homage to Otto III in 986, and
provided Polish armies for the campaigns. Saxon and Polish forces also coop-
erated against Bohemia. The Bohemian duke, Boleslav II in turn concluded an
alliance with the Liutizic confederation. Here tire conflicts between the Polish
Piasts and the Bohemian Przemyslids became visible; from about 990 the two
families fought for lordship over Silesia and Cracovia, and in the last resort for

The Ottomans as kings and emperors 2 5 7

hegemony over the whole area of west Slav settlement. A third family was also
concerned in diese rivalries, the Bohemian Slavnikids, who were of some
importance for Ottoman eastern policy and provided the bishop of Prague in
the person of tire subsequendy canonised Adalbert. The leadership in the wars
for the reconquest of the areas of Polabian setdement and the securing of the
marches of Meissen and Lausitz was taken by the archbishop of Magdeburg,
Giselher, who had worked for the absorption of the bishopric of Magdeburg
by Magdeburg in 9 8 1 , together with Margrave Ekkehard I of Meissen, who had
been appointed by Theophanu. All these efforts were in vain; the Liutizi and
Abodrites retained their independence and freedom.

This was not altered by the campaigns undertaken by Otto III after he
reached his majority: in die autumn of 995 togedier with the new Polish duke
Boleslav Chrobry; in the summer of 997, when the Liutizi threatened the
Arneburg on the Elbe and for a time even conquered it. After this Otto turned
away from this aspect of Saxon and Ottoman policy and towards Rome. His
policy of renewal of the Roman empire changed the angle of vision of
Ottoman policy, though it should be noted on the one hand that Rome and the
imperial position in Rome had played an important part in Ottoman policy
since 962 and on the other that the political and ecclesiastical developments in
the east did not lose their importance for Ottoman policy: die expansion of the
Polish realm of the Piasts; the menacing of the Czech realm of the
Przemyslids in Bohemia by Poland; the persecution of the Slavnikids in
Bohemia by the Przemyslids; the formation of a Hungarian kingdom under
the Arpads. Otto III had a particular link with missionary activities in these
areas through the bishop of Prague, Adalbert, who after being driven from his
bishopric sought a life as monk and missionary and was martyred by the
Prussians in 997.

In September 994 Otto III, now fifteen years old, received die arms which
marked his entry into manhood and hence his ability to rule in person. The new
king had received an education in Latin and Greek. He was imbued with a
belief in his divinely sanctioned and unrestricted imperial rule and widi an
enthusiastic religiosity directed towards asceticism and mission; through his
mother he inherited a particularly strong orientation towards Byzantium. This
was combined with an admiration for die example of Charles the Great and for
Carolingian tradition. These conceptions found expression in a programme of
renewal of the Roman empire, which Otto developed following his imperial
coronation in 996 under the influence of die chancellor Heribert, Leo of
Vercelli and Gerbert of Aurillac in particular. Gerbert, whom Otto made pope
in 999, addressed Otto as the new Constantine and as pope took the appropri-
ate name of Sylvester II. As emperor Otto combined the vision of a renovatio
with that of his apostolic status, which gave him die duty to see through an

258 ECKHARD MULLER-MERTENS

apostolic renewal of the church. On his first Italian expedition in 996, while
still king, he behaved as if the papacy were an imperial bishopric in appointing
his relative and chaplain Brun as pope, who as Gregory V became the first
German to hold the see of St Peter. Following the imperial coronation on 21
May 996, Otto refused either to renew the Ottonianum, the confirmation of the
Roman see’s privileges by Otto tire Great, or to acknowledge the Donation of
Constantine; he persisted in this attitude and indeed later rejected the latter
document as a forgery. He also claimed leadership within the church. He was
determined to exercise imperial rule over Rome and within the patrimonium
Petri ; the result was tensions with the Roman nobility and die curia. Although
Otto in effect conceded their wishes in die end, the concession of the eight dis-
puted counties in die Pentapolis injanuary 1001 took the form not of a recog-
nition of papal privileges but of an imperial donation.

The Christian renewal of the Roman empire was practised by Otto III and
his court after Otto had, in the course of his second Italian expedition, which
lasted from the end of 997 to the beginning of 1 000, renewed control of Rome
in February 998. At die beginning of this process stood the cruel tribunal over
the antipope John Philagathos who had been set up by the Crescentii, the exe-
cution of Crescentius II and die exiling of his supporters. The emperor had an
imperial palace built on the Palatine hill. Court tides derived from ancient
Rome and a court ceremonial on Byzantine lines were introduced. Like those
of the Byzantine emperors and the popes, the diplomata were validated by
metal bulls with the device renovatio imperii Romanorum (‘renewal of the empire
of the Romans’). The chancery had already adopted the tide of Emperor of
the Romans from die time of the imperial coronation. To diis Otto added the
apostolic devotion formula servus Jesu Christi (‘servant of Jesus Christ’) on his
journey to Poland, and from January 1001 the formula servus apostolorum
(‘servant of the aposdes’).

The most important actions with practical political consequences in the
course of the four years of die Renovatio were the diplomata and other acts of
Otto III and Sylvester II which reorganised imperial and papal connections
with Poland and Hungary. In Rome negotiations were conducted with repre-
sentatives of the Polish prince which led to the decision to found a Polish arch-
bishopric. This was carried out in die course of a pilgrimage which Otto III
undertook in February and March 1000 to Gnesen, where Boleslav Chrobry
had had the body of the martyred Adalbert buried. The emperor concluded a
pact of friendship with Boleslav. As a sign of his new status he set the imperial
diadem on his head, presented him with a copy of the Holy Lance, and turned
him, as Thietmar of Merseburg grumbled, ‘from a tribute-payer into a lord’d It

‘tributarium faciens dominum’: Thietmar, Chronicon v, io, p. 232.

The Ottomans as kings and emperors

259

is even conceivable that Otto made Boleslav king in a secular ceremony, though
if tiffs did occur it was not followed by the ecclesiastical consecration necessary
for the full legitimation of kingship. On the journey through the German terri-
tories which followed, Otto had tire crypt of Charles the Great in Aachen
opened and removed from it the emperor’s golden breast-cross. By die
summer of 1000 Otto was once again in Rome. The discontent within the
Roman nobility and at the curia against imperial rule grew; in the end a Roman
uprising forced emperor and pope to leave Rome in February 1001. Otto and
Sylvester took themselves off to Ravenna, where at Easter they gave their sanc-
tion in a synod to the reorganisation of the Hungarian church planned by
Stephen I and based on an archbishopric at Gran. Stephen, who like Miesco
presented his kingdom to St Peter and sought royal status, was sent a royal
crown by Otto III, the crown of St Stephen, with which he was crowned by the
new Hungarian metropolitan. An aristocratic conspiracy in Germany at this
time, about which we learn from Thietmar of Merseburg, may have had as its
driving force the injured rights and claims of the churches of Magdeburg,
Salzburg and Passau. News of its outbreak probably did not reach the emperor
before his death: he died on 24 January 1002, not yet twenty-two years of age,
near Rome, which he had not yet been able to reconquer. In the wake of the
cortege bringing his body to Aachen tire opponents of Ottoman rule in Italy
rose.

In the era of Otto III the Ottoman imperial structure and style of rulership
were developed, extended and intensified. There was an increase in the interac-
tion between royal court and the imperial churches and the use of the latter for
royal service. The absolute numbers of the royal chaplains, as well as of those
chaplains who were at the same time holders of canonries at cathedrals or
other foundations, increased substantially. Chaplains are found as royal mes-
sengers, participating in or presiding over the royal court, and also intervening
in royal diplomata. The most important political advisers of Otto III were
court chaplains and they remained permanent advisers of the king even after
they had become bishops or popes. After a pause between 984 and 989 the pro-
motion of chaplains to bishoprics became more and more frequent. The impe-
rial churches had their rights of immunity extended both in nature and in
extent. It was in the second half of tire tenth century that most churches had
their first grant of immunity coupled with bannus, the right to command and
to hold courts; there was a concentration of such grants in the 980s. These also
saw tire first grants of whole counties to bishoprics. From the 970s a fresh
wave of privileges granting annual markets, mint, toll and roads began, reach-
ing its high-water mark around die year 1000. Around 1000 Otto III also
issued the earliest privileges permitting weekly and daily markets. The counter-
part to these changes in the organisation of the kingdom was a new form of

z6o ECKHARD MULLER-MERTENS

legitimation of rulership: it was in these years that the ruler portraits so charac-
teristic of the late Ottoman and early Salian period first appear. The ruler,
though still alive, is here depicted as translated from his earthly surroundings
into the celestial sphere, projected on to the same level as Christ and tire saints.
The new type of portrait corresponded to developments within the Ottoman
theology of rulership. The king or emperor received his legitimation and his
promised position as Christ’s representative on earth through the service he
performed for God. The policy towards Rome was also marked by both conti-
nuity and intensification. The Ottomans had concentrated to a remarkable
extent on Rome and Italy since Otto’s departure for his imperial coronation.
Of the forty years and five months between August 961 and January 1002,
sixteen years and ten months had been spent on Italian expeditions. The
numerous Italian expeditions, which followed on one another in rapid succes-
sion, the lengthy courts held in Italy, the journeys across the Alps which these
entailed for German princes and their followings — mailed horsemen and other
fighting men as well as servants — and the journeys in the reverse direction for
Italian magnates all led to a lengthening and strengthening of the north-south
axis of integration of the Ottoman Reich, which now extended from the lands
around the Harz, through east Franconia into southern Germany and across
the Alps into Italy. For the first time in the Ottoman period Otto Ill’s reign saw
assemblies, meetings with magnates and the issuing of diplomata in Bavaria
and Suabia as the court moved to and from Italy. It was such practical matters
which helped to integrate Francia et Saxoniawith the south German duchies, the
remaining German territories and tire regnum Italicum. Emperorship, emperor
and tire whole empire came to be orientated towards Rome.

RENOVATIO REGNI FRANCORUM UNDER HENRY II: THE EMERGENCE OF
THE REGNUM TEUTONICUM AND THE DEVELOPMENT OF EMPERORSHIP

Otto III left no son, nor was there a surviving brother or brother of his father
who could succeed by hereditary right to the throne. There was only one sur-
viving member of the royal house in the male line: Henry IV of Bavaria, a
great-grandchild of King Henry I and the representative of the Bavarian line
of the Liudolfings. He at once made a bid for die crown. In addition there were
candidates in the persons of Duke Hermann II of Suabia, a Conradine, and
Margrave Ekkehard I of Meissen. Henry was supported by Archbishop
Willigis of Mainz, who had lost his previously dominant infiuence at court
under Otto III. On 6 or 7 June 1002 the Bavarian Liudolfing was elected as king
by his Bavarian, Frankish and upper Lotharingian supporters in Mainz and
anointed and crowned king by Willigis. By this time Ekkehard had already been
killed in a feud unconnected with the dispute over the succession. After an

The Ottomans as kings and emperors 261

indecisive campaign against Hermann of Suabia, Henry in die course of a per-
ambulation of the kingdom successively won the recognition of the
Thuringians, Saxons and lower Lodiaringians either by homage or by a
renewed election. Finally Hermann submitted in Bruchsal on 1 October 1002.
The new king did not retain Bavaria in his own hands, but he also did not grant
it, to the margrave of the Bavarian northern March, the Babenberger Henry of
Schweinfurt, as he had at first promised, but to one of the four brothers of his
wife Kunigunde, a member of the Luxemburg line of the clan of the Ardennes
counts. Henry of Schweinfurt, backed by Boleslav Chrobry, rose against this
decision and was crushed by Henry.

In Italy an anti-Ottonian group had set up Margrave Arduin of Ivrea as king
on 15 February 1002. As soon as Henry had established his kingship in
Germany he sent Duke Otto of Carinthia against Arduin around the turn of
1002/ 3. As early as the spring of 1004 he himself came to Italy, where on 14
May 1004 he was elected rex Langobardontm and crowned; between May 1004
and May 1005 he issued several diplomata under die tide of rex Francorum et
Langobardomm, thus acknowledging a certain independence of Italy from the
Frankish kingdom, his realm nordi of the Alps. Henry insisted on Burgundy’s
feudal dependence. He had himself acknowledged as heir by the childless King
Rudolf III in 1006 and had his overlordship confirmed on several occasions.
On 14 February 1014 Henry received imperial coronation in Rome. Benedict
VIII sought him out in 1020 in Bamberg to ask for armed assistance against
Byzantium. It was here that Henry II confirmed the Ottonianum of 962; he
responded to the pope’s appeal and led a powerful army against Byzantium in
southern Italy on his third expedition in 1021— 2. The three Italian expeditions
were short, lasting seven, three and nine months respectively, and the two stays
in Rome were measured in days. The emperor refrained from any interference
in Roman affairs, and within the Italian kingdom reverted to the policy of pre-
serving the balance of power as practised by Otto the Great.

The Bavarian Liudolfing thus broke with the orientation towards Rome, and
with tire intention of exercising practical and decisive imperial audiority over
and in Rome, both of which had been determinant features of Ottoman policy
in the previous forty years. Instead of the device renovatio imperii Romanorum
used by Otto III Henry adopted the formula renovatio regni Francorum (‘renewal
of the kingdom of the Franks’), which had already been used on the imperial
bulls of Louis die Pious, Charles III, Arnulf and Wido of Spoleto. Possibly it
had already been formulated under Charles the Great, and at all events it
expressed continuity with him. Henry was the first ruler to use the device as
king - between the beginning of 1003 and 1007 - and evidendy he announced
in this way his conception of imperial kingship with a concentration on the
regnum north of the Alps.

z6z ECKHARD MULLER-MERTENS

Henry made another U-turn in his eastern policy. From die end of the 970s
onwards the Ottomans had supported Polish expansion against the Przemyslid
principality, in the Baltic areas and against Kiev Rus', and received in return
armed help of the Polish dukes in their attempts to reconquer the lands of the
Liutizi and Abodrites. Then Otto III had made Boleslav Chrobry a member of
the imperium Romanum at Gnesen in 1 000. Boleslav was henceforth able to claim
to be ‘brother and cooperator of the empire, friend and ally of the Roman
people’. 9 In the course of die succession dispute of 1 002 he occupied die lower
Lausitz and the Milzener land around Bautzen following Ekkehard I of
Meissen’s violent deadi, presumably in agreement with the Ekkehardings.
Henry gave him these marches as benefices when Boleslav took part in the
renewed election by the Saxon princes in July 1 002 at Merseburg and with these
acknowledged the new king. At the beginning of 1003 the Polish duke seized
Bohemia. Henry demanded that he should do homage for die Bohemian
dukedom, which Boleslav refused. It was then drat Henry broke with previous
Ottoman policy: he renounced further attempts to reconquer the territories of
die Elbe Slavs lost in 983; at Easter 1003 he concluded an alliance with the
Liutizi and togedier with them took up arms against Boleslav Chrobry with the
intention of driving him out of Poland and out of die Saxon eastern marches
which had remained under German rule after 983. Like Otto Ill’s Roman and
Polish policy, Henry IPs alliance with the pagan Liutizi against die Christian
prince of Poland aroused violent criticism within Saxony. It can be found in the
chronicle of Thietmar, from 1009 bishop of the see of Merseburg restored by
Henry II in 1004, and in an admonitory letter written by the missionary arch-
bishop Brun of Querfurt in 1008. The Saxon nobility conducted the Polish
wars reluctantly and without endiusiasm. The Liutizi, who were themselves
interested in preserving a balance between the two opponents, did not contrib-
ute to a decisive victory by Henry. Henry and Boleslav conducted three cam-
paigns, in 1003-5, 1007-13 and 1015-18. The emperor was not able to deprive
the Piast prince of the lower Lausitz and die Milzener land: by the peace of
Bautzen of 1018 he confirmed the possession of these territories as benefices.
On the other hand he did prevent die incorporation of Bohemia into the Piast
empire and did preserve Ottoman overlordship over Bohemia.

Henry II remained within the paths and conceptions of Ottoman sacral
kingship. Having succeeded to the throne after his grandfather Henry had been
excluded from kingship in 936, he regarded himself as specially chosen by
God. His, die Henrician line of die Liudolfings, had been exalted by God after
a long period of trial and humiliation. This idea can be found in the younger
Vita Mathildis written between ioo2and 1012. With his sense of ruling as God’s

9 Gallus Anonymus, Chronicae et gesta ducum sive principitm Polonoinm , c. 6: ‘fratrem et cooperatorem

imperii constituit, et populi Romani amicum et socium appellavit’.

The Ottomans as kings and emperors

263

viceregent, Henry perceived all secular and ecclesiastical lordship as hierarchi-
cally subordinated to his kingship, and it was part of God’s order that it should
obey his power of command unconditionally and respect his authority in
matters of law and peace. Just as die Liuthar evangeliary of Aachen shows
Otto III as living ruler in earthly surroundings but directly in contact with
Christ and the divine sphere, so there is a portrait of Henry II in a Regensburg
sacramentary which does die same. Here the figure of Henry extends into the
nimbus around the figure of Christ, who with His left hand sets the crown on
the head of die praying king. Henry sought to shape the constitutional reality
according to such conceptions. He opposed private concentrations of lordship
and power and was concerned to stress the nature of aristocratic lordship as
office. In spite of this intensified Christian and sacral idea of kingship the king
remained, like all Ottomans, in fundamental consensus with the aristocratic
ducal and comital families. He completely respected dieir hereditary rights and
their expectation that die bishoprics and abbacies of royal monasteries would
be bestowed on members of the high nobility.

In the real world of rulership the king acted as umpire, judge and peace-
maker in the struggles for position which were produced by the formation of
regional lordships, in particular between the houses of the dynastic nobility
and between counts and dukes and the episcopal churches. Condicts in
Lodiaringia forced die king to undertake repeated campaigns between 1005
and 1012. Between die Meuse and the Scheldt the struggles turned around the
formation of what were later to become the territories Brabant, Hennegau,
Holland and Flanders, in particular between die Reginarids and the Ardennes
counts. In upper Lotharingia it was the Luxemburg branch of the Ardennes
comital house, related to Henry II by marriage, whose attempt to extend their
power by seizing the archbishopric of Trier in 1008 (the relationship had
already brought them the bishopric of Metz in 1005) set off a confiict which
lasted years. In Saxony too lay nobles sought to extend their power at die
expense of archbishops and bishops; as the king valued and favoured the latter,
tensions arose between him and the lay nobility which culminated in 1019-20
in a full-scale uprising by the Saxon duke and Saxon counts. Such struggles
were normally concluded by a compromise.

Henry II continued the policy of the Ottomans in integrating the imperial
churches - bishoprics, cathedral chapters and other canonries, and large mon-
asteries - into royal lordship and orienting them towards the court through
personal connections. The number of cathedral chapters linked with the royal
chapel rose, as did the number of royal chaplains promoted to bishoprics.
Between 1002 and 1021, twenty-two of the thirty- six episcopal vacancies
were filled by royal chaplains; twenty of the forty-seven bishops who died
during Henry’s reign were former chaplains. Henry was persistent and

264 ECKHARD MULLER-MERTENS

determined in having his candidates appointed. Occasionally he even
invested these against the will of or contrary to the proposal made by the
cathedral chapter; in Iris confirmations of privileges for bishoprics he occa-
sionally deleted the right of freedom of election granted in earlier privileges.
At the end of his reign his ability to get his own way may have been somewhat
reduced; almost all the bishops appointed in 1022 and 1023 were not
members of the royal chapel. The king-emperor encouraged the incipient ter-
ritories being built up by the imperial churches. He granted immunities with
ban and entire counties to bishoprics and royal monasteries. In all this
Henry’s church policy shows no real innovation, only development of exist-
ing tendencies, but the quantitative extension and intensification of the links
between court and imperial churches during his reign was so great as to give
the whole institution a new quality; Henry brought to fruition the specifically
Ottonian characteristics of the imperial church. Henry’s own contribution to
this is to be seen in the support he gave as king to the movement of monastic
reform, which established something like an ‘imperial monasticism’; already
in his period as duke he had supported monastic reform movements of
differing observances. He imposed Gorzian reform on the most important
imperial monasteries and reorganised both monastic life and the administra-
tion of monastic property, in some monasteries provoking by tiffs resistance
and even secessions by the monks. Possessions which exceeded the needs
implied by the Rule of St Benedict were taken over for the use of the
kingdom.

Henry II’s practice of government shows that as king he took over the
Ottonian positions in the lands around the Harz and based his lordship on
them. The political central regions in east Saxony/ north Thuringia, in the lands
around Rhine and Main, and in die lower Rhine/Meuse area continued to exist.
But the continuity was coupled with change. Henry extended the areas where
the court stayed for longer periods to include Alsace and die south German
duchies. At the same time he increased die speed of his itineracy and shortened
the length of his stays in the various parts of the Reich: these were visited for
shorter intervals more frequendy. Bavaria, Suabia and Alsace, previously
distant zones in relation to royal lordship, now became close zones.
Correspondingly the duchies of Bavaria and Suabia lost their special status.
They had continued to exist in the Ottonian Reich after 91 8 as duchies widi an
edinic basis, and so differed fundamentally from the duchies in upper and
lower Lotharingia and in Carinthia as well as from die Billung dukedom in
north-east Saxony. Henry II curbed the quasi-regal ducal lordship in Bavaria
and Suabia which he himself had exercised as duke, following the death of
Hermann II of Suabia in 1003 and with his own renunciation of Bavaria in
1 004. By founding the bishopric at Bamberg in 1 007 Henry created a new polit-

The Ottomans as kings and emperors

265

ical centre within tire Reich which lay on the north— south axis connecting the
Harz region via east Franconia with south Germany and northern Italy. This
reorganisation of government was carried out from Saxony, where Merseburg
played a special role as a place particularly favoured by Henry both to stay and
to hold assemblies at. The new bishopric at Bamberg received a substantial
landed endowment in order to fulfil its role as a centre of royal lordship, which
extended over the whole of the southern half of the kingdom and down along
important Alpine crossings.

Henry’s practice of rulership — adoption of the Ottoman central regions in
order to extend the practice of periodic royal presence over most of the
German territories, creation of a new centre of lordship in east Franconia in
the bishopric of Bamberg, curtailing of the viceregal dukedoms and other
prominent noble lordships, intensification of the use made of the imperial
churches for royal service and of their orientation towards the ruler’s court -
led to a new phase of integration of the kingdom, which now included Franks,
Lotharingians, Saxons, Bavarians and Suabians. The Roman and Italian policy
of the period from 961 to 1002 had given a new quality to the organisational
structures employed by rulership within the imperium in Germany and Italy. At
the end of tire Ottoman era both imperium and regnum were more integrated
than drey had been.

The emperorship of die Ottomans had already been linked with the Romans
by Otto II and Otto III. Contemporaries soon linked Henry II’s kingship with
the Germans. Around the turn of the millennium we find in Venice and south-
ern Italy the first references to a regnum Teutonicum and a rex Teutonicorum
(German kingdom, king of the Germans) applied to Henry II; his kingdom
north of the Alps, like that of Otto III, was a regnum Teutonicum. That contem-
poraries saw Henry II as a king of the Germans in his own lifetime is shown by
the diploma issued in 1020 for die Bavarian bishopric of Brixen near the
border with the Italian kingdom, which was drawn up by the recipient. The
bishop of Brixen’s scribe gave Henry the tide rex Teutonicorum, imperator augustus
Romanorum , 10 This tide remained an isolated instance during the eleventh
century, but it neverdieless clearly shows die processes of integration and con-
ceptions of reform of die Reich during the first decades of the new millen-
nium: the German and imperial components of the Ottoman Reich; die king
elected by the German princes with a claim to an emperorship of the Romans
and responsibility for the Roman empire, conceived of as simultaneously king
of die Germans and emperor of the Romans.

Emperor Henry II died childless on 1 3 July 1024 and was buried in the cathe-
dral at Bamberg. A cult soon grew up around Bamberg and die grave in the

10 ‘king of the Teutons, august emperor of the Romans’: D H II 424 .

266 ECKHARD MULLER-MERTENS

cathedral, culminating in the canonisation of Henry in 1 146 and of Kunigunde
in 1 200. Henry II was thus the first ruler in the succession to the Frankish
rulers to be perceived by contemporaries as king of the Germans, and die only
saint among the German kings and emperors.

CHAPTER IO

SAXONY AND THE ELBE SLAVS
IN THE TENTH CENTURY

Gerd Althoff

SAXON POLITICS IN THE SHADOW OF OTTONIAN KINGSHIP

Saxony after Carolinian incorporation

The conquest and incorporation of Saxony into the Carolingian empire, which
Charles the Great achieved after long and bitter struggles, had far-reaching
consequences for the political and institutional organisation of the Saxons.
The three Saxon ‘armies’ of the Ostfalians, Westfalians and Engrians, and the
‘national’ assembly of all castes - nobles, free men and freed men - ceased to
contribute to the coherence of Saxon political life. From 785 all assemblies in
Saxony were forbidden except for those summoned by a count or royal missus.
It was the so-called comital organisation which henceforth determined the
structure of lordship in Saxony; but no more fundamental Frankicisation of
the ruling strata in Saxony took place. The Saxon nobility allied itself with the
Franks, presumably by ties of marriage, and die Carolingian rulers did not
replace it with Frankish magnates. A second characteristic of the Carolingian
conquest was to have long-term consequences: Saxon territory did not become
a core region of the Carolingian kings, even after the divisions of the empire
among Louis the Pious’ sons. Carolingian visits to Saxony remained excep-
tional events. Already by the mid-nindi century we can observe nobles in
eastern and western Saxony termed dux. Between Rhine and Weser in die west
it is Ekbert, with his ducatus Westfalorum\ in the east it is the akvLiudolf, ancestor
of the Ottomans and founder of the nunnery of Gandersheim. In Saxony,
therefore, as in other parts of east Francia, we find that phenomenon known as
the ‘younger tribal duchy’: an aristocratic lordship claiming exclusive pre-
eminence widiin a gens (a source of lengthy feuds in Franconia and Suabia), and
consequendy bound to clash with die king. In Saxony such contiicts are not
recorded at first, and it is in any case doubtful whether the dux Liudolf already
claimed a leadership within the whole Saxon territory. He seems rather to have
confined himself to his own lordship in eastern Saxony, in the Harz mountains.

267

i68

GERD ALTHOFF

Liudolf’s position was inherited by his son Brun, who fell against the
Northmen in 880 while leading a Saxon army including the bishops of
Hildesheim and Minden among others. Liudolfing predominance was appar-
ently not affected by this, for Brun’s brother Otto ‘the Magnificent’ simply took
over his position. The influence exercised by tire Liudolfings is seen not least in
the way their women married east Frankish Carolingians: Liutgard married
Louis the Younger and Oda Zwentibald. Multiple alliances with the royal
family were accompanied by marriage links with powerful noble families outside
Saxony, for example with the Babenberger of eastern Franconia, but not
apparently by marriages among the Saxon nobility itself. Several of Liudolf’s
daughters remained unmarried as abbesses of Gandersheim, a very clear indi-
cation of the Liudolfings’ prominent position.

Henry, son of Otto and later to become king, was the first to break with
these marriage practices. He first married a widow from the Merseburg regions
for her rich inheritance, and then fell in love ‘because of her beauty and her
wealth’, 1 with the young Matilda, a descendant of the Saxon duke Widukind,
whose lands lay mainly in western Saxony. Ecclesiastical protests against the
first marriage made possible its dissolution and Henry’s remarriage; linked with
it was a thrust into western Saxony, which soon brought Henry into conflict
with the Conradines. The Liudolfings’ position was consolidated by Henry’s
successful conduct of his dispute with King Conrad I, who had at first tried to
restrict his succession to his father after the latter’s death in 91 2. Henry’s rise is
all the more remarkable when we consider that no other Saxon aristocratic
group succeeded in continuously extending its position in this way: neither the
descendants of tire dux Ekbert nor those of the Saxon leaders Widukind and
Hessi achieved such a continuity, though the reasons are not clear.

However, one very forcible reservation must be made when considering
Saxony in the late ninth and early tenth century: any judgements are made very
difficult by the fact that many evaluations come from the later ‘Ottoman’
historiography, which bathes early Liudolfing history in a flood of transfigur-
ing light. This is true not only of the works written in Gandersheim but also of
Widukind of Corvey’s Saxon History. It is thus ultimately unclear how the
Franks and the Saxons came to elect the Saxon duke Henry as king in 919 fol-
lowing King Conrad’s death.

The Saxon dukedom in the tenth century

Henry I’s elevation to kingship at Fritzlar in 919 had a less-noticed conse-
quence for the Saxon people: its duke was now king. Henry I is not known to

1 Thietmar, Chronicon , i, 9 .

Saxony and the Elbe Slavs in the tenth century

269

have taken any steps to install a substitute as duke in Saxony, and nor do the
Saxons themselves appear to have been active in this direction. This was the
start of a development which had great significance for the tenth century The
king, frequently absent, was able to shape Saxon politics less continuously than
would have been possible for a duke. He nevertheless refrained from establish-
ing an ‘office-like’ dukedom there, by contrast with the position in the south
German duchies. This had three characteristic consequences for Saxon history
in dais period: noble families forced their way into the resulting power vacuum;
it became more difficult to reach an expression of political will at royal and
popular assemblies, for a number of different forces had or claimed positions
of rough equality; cooperative forms of association (coniurationes) were wide-
spread, which is probably also a result of the organisation of lordship within
Saxony.

The Ottoman rulers naturally could not avoid giving offices and tasks to
Saxon nobles from which a pre-eminence within the people might have been
derived. The best-known such example was the appointment of Hermann
Billung as princeps militiae by Otto I in 93 6. 2 This had particularly important con-
sequences because with time Hermann’s position gradually broadened into
that of a duke, a position which became heritable within his family. But Otto’s
decision is also of great interest because it lit the fuse for a whole series of con-
flicts. These show that even in the early tenth century Saxon nobles drought of
themselves as having claims to particular offices and were not prepared to
accept royal decisions they perceived as arbitrary. Otto I’s choice was a - prob-
ably deliberate - breach of the internal ranking of the Billung family and of the
Saxon nobility as a whole, intended to demonstrate his right to make such deci-
sions. Resistance articulated itself immediately: Hermann’s older brother
Wichmann left the royal army and allied himself with Otto’s enemies, while an
Ekkehard, ‘son of Liudolf’, tried by an act of hare-brained courage to show
that he was a more appropriate choice for military leadership dian Hermann.
With eighteen companions he risked an attack on the enemy against royal
orders, and perished widi all his comrades. 3 In the following period we can see
how Otto I repeatedly entrusted his princeps militiae with representing him (pro -
curatio) in Saxony when he went to Italy. These procurationes came to be quite
lengthy ones, for die emperor spent almost all his time in Italy after 961. It is
therefore significant that the royal chancery refers to Hermann only as comes or
marchio , while Widukind of Corvey terms him dux, though he also uses this
term for other Saxon nobles of the period, such as Margraves Gero and
Dietrich. In spite of this it is clear that Hermann’s exalted position was also
reflected in his titles.

2 Widukind, Res gestae Saxonicae 11,4. 3 Ibid. 11, 4.

270

GERD ALTHOFF

The independence with which he acted can be seen from two events from
his second procuratio. Otto wrote to the duces Hermann and Dietrich, ordering
them not to make peace with die Redarii; although the letter was read out at a
popular assembly in Werla, the peace, already concluded, was kept, since the
Saxons feared to divide their forces while threatened by a war with the Danes. 4
Even more striking than this overriding of orders from a distant emperor
determined by die military position was a reception which Hermann arranged
for himself in Magdeburg in 972. Here he usurped the reception ceremony
reserved for die king, sat in his place and slept in his bed. He had, of course,
assistance in this, notably from the archbishop of Magdeburg, Adalbert, whom
Otto condemned to a hefty fine for his presumption. 3

It is not until the time of Bernard I, however, who had inherited all his
father’s rights in 973, that we learn that the position of the Billungs within
Saxony had come to approach that of a duke. In the crisis period after the
death of Otto II, who had left only his three-year-old son Otto III, just
crowned as co-ruler in Aachen, to succeed him, decisions by the Saxon people
were particularly necessary because Otto’s nearest male relative and guardian,
Henry the Quarrelsome, himself aspired to kingship and sought a decision in
Saxony. He summoned a Saxon assembly for Palm Sunday 984, and suggested
to it that he should be elected king. Some of those present expressed reserva-
tions, claiming to be unable to act without the permission ( licentia ) of Otto III.
We can tell drat Duke Bernard was among them, even though this is not
recorded explicidy, from the fact that he headed the participants at an assembly
held at the Asselburg immediately afterwards, intended to rally those who
wanted to resist Henry’s usurpation. 6 Although several Saxon counts besides
Bernard took part in this meeting, they were evidendy not strong enough to
prevent Henry from being proclaimed king by his supporters at the Easter cel-
ebrations of 984 in Quedlinburg. These supporters included the dukes
Boleslav of Bohemia and Miesco of Poland and the Abodrite prince Mistui. So
we cannot deduce an established ducal position for Bernard from these events,
even though he is named among Henry the Quarrelsome’s principal oppo-
nents.

The Saxon opposition to Henry, in alliance with forces outside Saxony,
notably Archbishop Willigis of Mainz, ultimately forced Henry to abandon his
plans for kingship and affirmed Otto Ill’s succession under the regency of his
mother Theophanu, while Henry was restored to the duchy of Bavaria. The
reconciliation and compromise were celebrated and publicly demonstrated at
the Easter celebrations at Quedlinburg in 986, certainly not by coincidence.
Four dukes served at table, as if participating in the coronation banquet of a

4 Ibid, in, 70. 5 Thietmar, Chronicon it, ib. 6 find, tv, 2.

Saxony and the Elbe Slavs in the tenth century 271

newly crowned king: Henry of Bavaria as steward, Conrad of Suabia as cham-
berlain, Henry the Younger of Carinthia as buder, and Bernard I as marshal. 7
When four dukes had participated in the coronation feast of Otto I in 936
there had been no Saxon duke present, and Widukind, to whom we owe the
description of the scene, talks of Siegfried in this connection as tire ‘best of
the Saxons and second after the king’, not as dux. s This may serve to underline
the Billung’s growing into a new role.

The royal succession following Otto Ill’s death also proved difficult; it is our
next opportunity to observe whether Bernard exercised ducal functions. First
he favoured his relative Ekkehard of Meissen, which made him a member of a
group ultimately unable to prevail even within Saxony. It was this partisanship
which perhaps explains why Bernard played no specially prominent role in the
formation of a Saxon view, for another Saxon, Margrave Liuthar from the
Walbeck family, was able to raise his profile as an opponent of Ekkehard of
Meissen. The Bavarian claimant to the throne, the later Henry II, even turned
in letters to his cousins, the Ottoman abbesses Adelaide and Sophie, in order to
win them and the Saxon magnates for his cause, with decisive effect. Details
like this show what varied forces could operate within Saxony alongside and
also against the duke.

Bernard’s partisanship for Ekkehard evidently did not damage him. At
Henry II’s so-called ‘subsidiary election’ ( Nachwahl ) in Merseburg, which
brought Saxon recognition of the new king, tire duke appeared as a representa-
tive of tire people before Henry II, who had appeared in full regalia, set out the
Saxons’ view of the matter and their own position, and demanded binding
promises. After Henry had given these, ‘Duke Bernard took up the Holy Lance
and entrusted him with the care of the Reich in tire name of all’. 9 The scene has
become a locus classicus for the scholarly investigation of the Saxon dukedom:
the Saxon duke can here be seen transformed from a ‘king’s representative
among the Saxon people’ to a ‘representative of the Saxon people before the
king’. 10 However, this is die only occasion when we can point to a Saxon duke’s
taking an important position at tire head of the people. It should also be noted
that the later Ottoman rulers no longer entrusted the Billung dukes with repre-
senting diem in Saxony during their absences in Italy, preferring instead to
make use of people who stood closer to them. Otto III chose his aunt Matilda,
abbess of Quedlinburg, famous for the care with which she presided over the
Saxon assemblies, and Henry II used the archbishop of Magdeburg and his
wife Kunigunde for the purpose. Such details should make us drink hard when
asked to determine die essential nature of the Billung dukedom. Signs of a

7 Ibid, iv, 9. 8 Widukind, Res gestae Saxonicae n, 2.

9 ‘Bernhardus igitur dux, accepta in manibus sacra lancea, ex parte omnium regni curam illi fideliter

committit’: Thietmar, Cbronicon v, 1 6— 1 7. 10 Jordan (1 9 5 8), p. 8.

272

GERD ALTHOFF

steady evolution of die Billungs into dukes are matched by others which cast
doubt on whedier kings, Billungs and odier forces within the Saxon nobility
shared a consensus about the rights and duties of the Saxon duke. There is thus
every reason for modern scholarship to rethink the characteristics of ducal
lordship in die east Frankish/German kingdom . 11

Observations on the Saxon nobility in the tenth century

It may be dynastic coincidence or not, but apart from the Liudolfings we know
of no Saxon noble family whose genealogical connections can be traced with
certainty from die ninth through to the tenth century. This is linked with
another observation. In ninth-century Saxony we can make out five noble
groups: the Liudolfings, die Ekbertines, the descendants of Widukind, the
Hessi-clan and the so-called older Billungs. In the tenth century, by contrast,
we can make out more than twenty-five possessing some sort of firm geneal-
ogy and visible history. Such a comparison suggests fairly certainly that the
Saxon nobility’s lordship profited to a considerable extent from the
Liudolfings’ rise to kingship and from die Ottoman dynasty. The connection
becomes still clearer if we ask whereabouts in Saxony the families visible only
from the tenth century onwards resided: for the most part they came from the
Liudolfing and Ottoman core region around the Harz mountains. This was
the home of the families associated with Margraves Siegfried and Gero, of the
Ekkehardines, Walbecks, Haldenslebens, of the counts of Weimar, Northeim
and Katlenburg as well as others, and if one notes that it was also here that the
well-known Ottoman palaces and house monasteries like Quedlinburg,
Gandersheim and Nordhausen lay and that Otto had by founding Magdeburg
and its suffragans of Merseburg, Zeitz and Meissen installed further ecclesias-
tical centres in the region, then one gets a picture of a concentration of lord-
ship unparalleled in die rest of the Ottoman Reich. Here we should note that
the borders between Saxony and Thuringia seem to have been permeable.
There are clear signs of a specific consciousness among both Thuringians and
Saxons, culminating in the election of Ekkehard of Meissen as duke of
Thuringia, reported by Thietmar of Merseburg , 12 and the demonstrative visit
by Henry II to Thuringia in the course of his itinerary round the Reich follow-
ing the royal election in Mainz. But on die odier hand Thuringian magnates like
the Ekkehardines or the counts of Weimar appeared at Saxon popular and
royal assemblies and cannot be distinguished from the Saxon magnates diere.
It is more plausible to speak of a Thuringian special consciousness within
Saxony.

11 See now Becher (1996). 12 Thietmar, Chronicotrv, 7.

Saxony and the Elbe Slavs in the tenth century 273

Similar concentrations of lordship are not found in the other regions of
Saxony outside this east Saxon/Thuringian core region. In the tenth-century
north two noble groupings, the Billungs and the counts of Stade, achieved
prominence and shared the military command on the Slav border. Even this
cannot be paralleled in western Saxony, in the bishoprics of Minden,
Osnabriick, Paderborn and Munster, apart from some evidence for lordships
held by the wider Billung family in these regions. It thus seems reasonable to
suppose that the concentration of noble families in the east Saxon/Thuringian
region, the Ottoman heardand, is connected with the patronage of these fami-
lies by tire royal dynasty. It has also been noted that in the tenth century royal
gifts of land to die nobility lay in the area where the recipient held office, in
sharp contrast to Carolingian practice. In other words, the gifts favoured the
development of a centre of lordship, and so in the long term the emergence of
‘territorial lordships’ held by nobles and prelates. Indirect evidence for such
concentrations around a centre of lordship is also provided by the numerous
monastic foundations of the tenth-century nobility. These were mausolea,
centres preserving the memoria of the family’s relatives, and so providing points
of crystallisation around which noble family consciousness could be preserved
and cultivated. The cultivation of memoria also allows us to see the very close
ties between individual noble families. In the course of the tenth century the
Saxon nobility had become quite exceptionally interrelated, even if dynastic
marriage for political reasons is hardly a Saxon invention. It is noteworthy
how closely the most eminent Saxon families were related to each other: the
Billungs to the counts of Stade, the family of Margrave Gero and
the Ekkehardines; the counts of Stade in turn to the counts of Walbeck; die
Ekkehardines also to the counts of Walbeck. This list, owing much to die gene-
alogical information provided by Thietmar of Merseburg, could easily be
extended, but it can hardly be doubted that the most influential sections of the
Saxon nobility in the Ottoman era were all related to one another. This did any-
thing but prevent conflicts — one might offer examples from the history of the
Billungs and of others - but it also certainly offered chances for the members
of this stratum to act together and to exchange information, very necessary at
royal and popular assemblies but also for the successful conclusion of coniura-
tiones.

Problems of political consensus-formation in Saxony: popular assemblies and
coniurationes

Saxony in die Ottoman era was rich in centres from which royal lordship was
exercised. For this reason it is clear that the politics of the gens were to a large
extent decided at such places, for example at die Easter palace at Quedlinburg

2 74

GERD ALTHOFF

or the urbs regia Magdeburg. Alongside such centres, however, Werla grew in
importance as a location for Saxon popular assemblies. It must be stressed that
the popular assemblies there of the Ottoman period all took place without
kings: the generally held view that tenth-century kings often appeared at these
assemblies has no foundation. Werla was certainly an urbs regia, and the
Ottomans frequently stayed there, but their stops were spread across the year
and do not suggest regular participation at a regional popular assembly in
springtime. It is, however, very questionable whether there was aj yearly popular
assembly at all. Three of the four pieces of relevant evidence come from
periods when there were problems with royal succession following the deaths
of Otto II, Otto III and Henry II in 984, 1002 and 1024 respectively. The
fourth piece of evidence comes from 968, when the Saxons met in Otto I’s
absence in Italy under the leadership of their duces Hermann Billung and
Dietrich.

These instances also show that it was not the duke who summoned these
assemblies; rather, the cooperative forms of organisation among the Saxon
magnates brought about the meetings at Werla. In 984 Henry the Quarrelsome
hurried to Werla in order to prevent or pacify a coniuratio which was taking place
there. In 1002 Saxon magnates pledged themselves at a secretum colloquium in
Frohse that they would elect no one as king before a meeting to take place at
Werla. Thietmar of Merseburg gives us a very precise account of the course of
this colloquium , showing the form such an assembly took. The claimant Henry
sent a messenger to the assembly, in particular to the Ottoman abbesses Sophie
of Gandersheim and Adelaide of Quedlinburg, as well as to the Saxon mag-
nates, showing the important role played by Ottoman princesses in such meet-
ings. At the meeting die messenger revealed his master’s offer to reward
generously all who would help him to kingship. Thietmar shows that by no
means all politically relevant forces participated, for Margrave Ekkehard of
Meissen and his following were absent, and it is pretty certain drat this included
Duke Bernard and the bishops Arnulf of Halberstadt and Bernward of
Hildesheim. Ekkehard, together with Bernard and Arnulf, provoked the
Ottoman ladies and their guests by usurping a festive banquet and eating up the
meal probably intended in favour of those who had decided for Henry as
king. 13

The active role played by female members of the Ottoman family in Saxon
politics is also evident in die succession of Conrad II in 1024-5. The Saxons
again organised an assembly in Werla to discuss the royal election and other
outstanding problems, including a reconciliation between Bishop Meinwerk of
Paderborn and the Billung count Thietmar. But the main effect of the assembly

13 Ibid, v, 4.

Saxony and the Elbe Slavs in the tenth century 27 5

was that the Saxons took no part in die election of Conrad II at Kamba; yet
there was, in contrast to 1002, no longer talk of a native Saxon candidate. When
Conrad then proceeded to Saxony via Lotharingia the Ottoman abbesses
Adelaide and Sophie behaved differently from die other Saxon magnates, who
did homage to tire king in Dortmund and Minden. The ‘imperial daughters and
sisters’, as they are termed in tire Quedlinburg annals, received the new king,
their distant relative, with demonstrative affection as the ius consanguineum pre-
scribes, in Vreden in western Saxony. 14 It is evident that in so doing they largely
predetermined the outcome of the Saxons’ decision.

All these details show that in Saxony there were traditional forms of reach-
ing political decisions, held for preference in Werla but also in other places like
Frohse, Seesen or the Asselburg. Equally clear is the fact that the Saxon duke
had no pre-eminent role in die summoning and conduct of these assemblies; it
is rather the variety of political forces operating in Saxony which is noteworthy,
including the female members of the Ottoman house in a prominent position.
It was not only Sophie and Adelaide who were to be found in the first row at
such assemblies; Matilda of Quedlinburg - as already mentioned - is also
praised for her careful presidency over them during Otto Ill’s absence, ‘sitting
amongst die bishops and dukes’. 15 But it was not only members of the royal
family who might take a higher profile than the dukes at these assemblies.
Margrave Liuthar clearly dominated an assembly at Frohse in 1002; he sum-
moned the ‘better part’ of the assembly from a consultation to a secretum collo-
quium. , after the larger assembly had failed to reach consensus. The confidential
meeting allowed die controversy to be worked dirough; it had arisen over die
aspirations of Ekkehard of Meissen to kingship and culminated in the well-
known piece of dialogue: ‘What have you got against me, Count Liuthar?’
‘Can’t you see that your wagon is missing a fourth wheel?’ 16 These details also
reveal the cooperative forms and structures of such meetings, which do not fit
with a model of a popular assembly directed by the duke. It would seem diat
the lack of continuity in the duke’s position on the one hand and the sense of
dignity held by margraves, archbishops, bishops and royal abbesses on die
other hand both helped to shape political structures within the Saxon people.
Since ducal lordship was inadequately established, the politically active forces
made use of forms of communication and interaction customary among
cooperatively structured groups. This helps to explain why oath-takings played
such an important role at these assemblies. Thietmar of Merseburg calls the
arrangements made by Henry the Quarrelsome’s opponents at a meeting in
Werla in 984 a coniuratio. Henry took it so seriously that he sent a bishop to
negotiate with it, who fixed a date for a meeting. This coniuratio had already been

14 ‘imperiales filiae ac sorores '\A.nnales Quedlinburgenses, s.a. 1002, p. 78.

15 Annales Quedlinburgenses, s.a. 999, p. 75. 16 Thietmar, Chroniconw, 52.

276

GERD ALTHOFF

prepared at a meeting at the Asselburg; Thietmar names those who took part,
calls them consocii and their actions conspirare , 17 It was also an oath which bound
the participants at tire meeting in Frohse to act in common at Werla in choos-
ing a king; they too were a sworn confederation.

These forms of cooperative coniuratio associated with the royal elections of
the late tenth century remind us that in early tenth-century Saxony magnate
coniurationes stood at die beginnings of actions against the king. For the conclu-
sion of such sworn confederations there was a traditional location, Saalfeld in
Thuringia. The coniuratio into which Henry, Otto I’s brother, entered with
Saxon magnates in 939 and Liudolf, Otto’s son, in 951, was marked by a conviv-
ium , a festive banquet intended to create and strengthen community and gener-
ally characteristic of such cooperative associations. Members of the royal
house initiated sworn confederacies, in other words; drey joined Saxon nobles
in taking oaths. Contemporary historians claimed that observers regarded the
very existence of such meetings as sinister, for their purpose was evidendy
unambiguous: people met to enter into a coniuratio at a specific place, confirmed
it with a banquet and then hastened to turn their oath into reality. In both of
the cases just mentioned this meant beginning a feud against Otto the Great.
The possibility of summoning socii and coniuratores to assistance in seeking
revenge for any injustice suffered strengthened, as will easily be seen, the posi-
tion of all members of such coniurationes. Even though such associations need
have had no institutionalised continuity, we should not underestimate the per-
manence of the links they created. The circle of participants in tenth-century
conspiracies against the king leaves an impression of remarkable constancy.
After all, Tacitus had already pointed out drat among the Germans both amici-
tiae and inimicitiae were inherited from father to son. This was undoubtedly
strengthened by the fact that cooperative ties were joined by familial ones. It is,
of course, not a Saxon peculiarity, but we should once again stress how closely
related all politically relevant forces in Saxony were; it can be said without exag-
geration that all were locked into this network of cooperative and familial ties.
Admittedly, this did not prevent conflicts, but it did create or facilitate the pos-
sibility of regulating them, because comrades and relatives were familiar with
the idea of ‘compensation through satisfaction’. How this worked can be seen
from the example of the members of the Billung family Wichmann and
Ekbert, the ‘classic’ rebels of the Ottoman era. Following their feuds and
actions against king or duke, they repeatedly found mediators and advocates,
who secured their re-entry into die king’s grace or provided them with an
opportunity to flee. This ‘system’ of conflict resolution was so familiar in
tenth-century Saxony that in 984 Henry the Quarrelsome’s cause was severely

17 Ibid iv, 2 .

Saxony and the Elbe Slavs in the tenth century

2 77

damaged by his behaviour towards two Saxon counts, Dietrich and Siegbert: he
spurned them as they begged forgiveness, barefooted, for earlier offences.
They responded appropriately by persuading their friends and relatives to quit
Henry’s following.

Overall, ties of family and sworn association made a substantial contribu-
tion in tenth-century Saxony to the stabilisation of political relationships. This
is certainly due to the fact that the structures of lordship within the gens were
underdeveloped, there being no clearly defined ducal position. The Ottoman
kings from Saxony provided the people with an enhanced sense of being an
‘imperial people’, but such a consciousness had little effect on Saxony’s internal
political structures. This situation naturally became highly problematic as soon
as Saxony ceased to provide the ruler, as happened in 1 024 and already in effect
in 1002, when Henry II, a Bavarian Liudolfing, ascended the throne. His rela-
tionship with the political forces in Saxony was fraught with problems.
Substantial parts of die Saxon nobility and episcopate were united in rejecting
Henry’s policy towards the Elbe Slavs and Poland, and this led to more or less
open conflicts diroughout his reign. Henry’s measures towards ‘centralisation
of governmental power in the Reich’ thus inevitably met with severe con-
straints in Saxony. 18 It is symptomatic of this that in 1020 Duke Bernard II
‘moved all Saxony with him in rebellion against the king’. 19 Bernard’s rebellion
had no consequences for him, for Archbishop Unwan of Hamburg-Bremen
and Henry’s wife Kunigunde were able to mediate a peace without any diminu-
tion of the Billungs’ possessions and rights. Once more we find the character-
istics of Ottoman ruling practice, which owed more to mediation and royal
clemency than the assertion of claims to power. But diis already prefigures the
severe conflicts between the Saxons and the Salian rulers, in which die claims
of the Salians to lordship and the quite different customs of the Saxons clashed
openly. This clash between hierarchical and cooperative principles and struc-
tures of lordship culminated in the ‘civil war’ under Henry IV and Henry V, but
these kings remained unable to destroy the established organisational forms of
the Saxons. In the Saxon wars under the Salians we again hear of the conventicula
and colloquia of secular and ecclesiastical magnates, which led to coniurationes, in
other words of cooperative forms of organisation, widiout die duke’s playing a
decisive role. Henry the Lion was ultimately to fail in die face of this inheri-
tance.

18 See the title of the article by Weinfurter (1986).

19 ‘totam secum ad rebellandum cesari movit Saxoniam’: Adam of Bremen, Gesta 11, 48.

2 7 S

GERD ALTHOFF

SAXON RELATIONS WITH THE ELBE SLAVS

Conquest and incorporation ?

For a long time German historiography depicted the politics practised by the
Ottomans and Saxons towards their Slav neighbours as a particularly dynamic
aspect of the ‘German eastward movement’, while the historians of the Slav
states took the same subject as the basis for a diagnosis of an aggressive
‘German drive to the east’. For the Ottoman period it was established that the
kings had had a plan of expansion which aimed at extending the Reich at least
as far as the Oder, in other words to incorporate die coundess small Slav
peoples between Elbe and Oder or Neisse. Scholars assumed that these expan-
sionary plans were accompanied by a plan of missionary activity, as developed
and realised by Otto the Great in particular, with his foundation of an arch-
bishopric at Magdeburg and of suffragan bishoprics at Meissen, Zeitz,
Merseburg, Brandenburg and Flavelb erg. Just as Charles the Great simultane-
ously Christianised the Saxons and incorporated them into his empire, so the
Ottomans and their helpers are supposed to have intended to act similarly
towards the Elbe Slavs, perhaps even towards die already Christianised Poland
and Bohemia. By contrast with Charles the Great, however, the plans of the
tenth century met with litde success.

As early as die nineteenth century German historiography had sought and
found an explanation for this in die rulers’ Italian policies, which had absorbed
essential energies of both the German Reich and its rulers and distracted them
from the national tasks awaiting them in the east. It is certainly true that the
results of both expansion and mission were only modest. The well-known Slav
uprising of 983 largely destroyed what had been established before that date;
the episcopal sees of Brandenburg and Havelberg remained orphaned until the
twelfth century. Only in the south did Ottoman lordship in the Slav regions
have some permanence. Astonishingly, succeeding kings did not make any
serious effort to wipe out this disgrace. In spite of a few military actions it may
be doubted whether the last Ottoman rulers even had such a recuperation as
their primary goal. Otto III enhanced the position of both the Polish and the
Hungarian rulers through some spectacular actions, and was prepared to
concede them an independent church organisation by helping with the foun-
dation of their own archbishoprics. Henry II by contrast concluded an alliance
with die pagan Liutizi in 1003 against the Polish ruler, probably entailing a
renunciation of missionary activity. 21 The last two Ottoman rulers thus did not
orientate themselves according to the plans which scholars have assumed to

20 Brun of Querfurt, Epistola ad Heinricum regem , ed. Karwasirika, pp. ioi ff.

Saxony and the Elbe Slavs in the tenth century 279

have motivated their predecessors, and in consequence contributed not a little
to the poor balance-sheet - seen from a nationalistic perspective - of the
Ottomans’ eastern policy.

This in sum is the dominant view of Ottoman eastern policy. Since the 1 960s
there have been strenuous efforts to write about the situation along the
German-Slav borders as a history of relations, to shift attention away from
confrontation towards varied exchange and a neighbourly living together. This
reorientation was important and justified, but it concentrated largely on other
themes of German-Slav history; the older view of Ottoman eastward expan-
sion was not explicitly replaced by a new assessment. One may well ask,
however, whether it really corresponds to the political conceptions of the
tenth century to see expansion, extending of borders and incorporation of the
Elbe Slavs, perhaps even the Poles, as a major goal of Ottoman and Saxon
policy. If that had really been so, we should have to ask why die military might
of the European hegemonial power was unable to conquer and incorporate
the small Slav tribes along the eastern border. One answer is readily available: it
was not brought to bear. Although this may seem at first sight surprising, the
Ottoman rulers did indeed refrain from applying their Reich’s military power
and potential to eastward expansion. ‘Imperial expeditions’ eastward, by
analogy with Italian expeditions in which forces from the whole Reich partici-
pated, did not exist in the Ottoman period, apart from Henry II’s wars against
the Poles. Military conflict widi the Elbe Slavs was essentially a matter for the
Saxons, and as far as we can see primarily for the Saxons living along the
borders. Ottoman rulers from Henry I to Otto III certainly participated in
campaigns against the Slavs, but leading Saxon contingents. Even when the
Slav uprising of 983 brought the ‘bold arch of Ottoman eastern policy’, in
Briiske’s phrase, so to speak to the point of collapse, it was an exclusively
Saxon contingent under die leadership of the margraves and of the archbishop
of Magdeburg which met the insurgents. 21 In the following decades there is
litde to be seen of military attempts to reconquer the areas lost in 983 .

If the ineffectiveness of the alleged expansion already suggests doubts
about its putative dynamism, these are strengthened once one asks how con-
quest and incorporation are to be envisaged in concrete terms. There is no
echo of the Carolingian model of an exchange of elites based on a newly intro-
duced comital system. By contrast with the behaviour of Carolingian armies in
Saxony, Saxon contingents never remained for long stretches in the lands of
the Elbe Slavs; they returned home successful or unsuccessful, a pattern which
hardly suggests an intention to expand permanently.

What expansionary plans can in any case be ascribed to an Ottoman king-

21 Briiske (1983), p. 36.

i8o

GERD ALTHOFF

ship which within the Reich largely did without an administration of its own,
and in many respects contented itself with an acknowledgement in principle
of its overlordship or primacy, without drawing concrete advantage from
such an acknowledgement? Certainly, Ottoman and Saxon eastern policy
aimed at recognition of its overlordship by the Slav peoples, which was to
find expression in regular tribute-payments. Certainly, Otto the Great under-
took considerable efforts to create the preconditions for successful evangel-
isation of the Elbe Slavs, through founding and endowing bishoprics. Had
these peoples become substantially Christianised, their relations with the
Ottoman Reich would undoubtedly have had to be rethought, as die exam-
ples of Bohemia and Poland show. But conquest and incorporation are
notions which characterise the aims and goals of warfare in the early modern
or modern period. Their transfer to the political and military conflicts of the
medieval period, especially of die tendi century, is distincdy problematic:
expanding and conquering rulers like Charles the Great were die exception
rather than the rule. Whedier Otto I of all people had set himself the task of
building up a ‘centrally directed and firmly organised and effective administra-
tion’ in die lands of the Elbe Slavs is somediing which has to be proved in
detail and not merely assumed, 22 not least because such an assumption is
difficult to sustain, given the other limiting conditions affecting Ottoman
lordship.

If we repeatedly hear about campaigns, surrenders, payments of tribute,
submissions and recognition of overlordship, followed by renewed conflicts
and submissions, widiout our being able to detect any steps towards more per-
manent incorporation, then we must ask whether die basic aim of Ottoman
rulers was not precisely to secure this recognition of overlordship, though
there was certainly also a missionary aim, pursued with varying degrees of
intensity. The dubious role played by the medieval eastern policy of Germany
in die political arguments and practice of the twentieth century, from the
Weimar Republic through the Third Reich to the period after the Second
World War, makes it essential to ask whether the situation along the eastern
border in the tenth century has not been fundamentally misinterpreted. In
what follows, the sources - not in any case numerous - will be interrogated
again to see what they actually have to report about relations between Saxons
and Elbe Slavs. The aim is to show that contemporary reports describe the
conflicts using categories which leave litde room for concepts like expansion
and incorporation.

22 Ludat (1968/1982), p. 46.

Saxony and the Elbe Slavs in the tenth century

281

Border warfare: submission, tribute, reprisals, peace

The situation along the border between Saxons and Elbe Slavs, which had
always been characterised by plundering expeditions on both sides, changed
after Henry I became king for a specific reason: the Slavs east of the Elbe
served as a testing ground for the troops, in particular mounted soldiers, whom
Henry was training for a defence against die Hungarians following his nine
years’ truce with them in 926. The years 928 and 929 saw a campaign by Henry
against the Hevelli with die conquest of die Brandenburg, another against the
Daleminzi followed by die erection of the fortification at Meissen, a counter-
attack by the Redarii leading to the destruction of the fortification at
Walsleben, the destruction of a Slav army in the batde of Lenzen under the
leadership of the ‘legate’ Bernard and his ‘colleague’ Thietmar, and finally a
campaign by Henry against the Bohemians. According to Widukind, 23 Henry I
did not repudiate the truce with the Hungarians until he had sufficient warriors
trained in mounted combat.

That such confiicts followed different principles from those customary in
the feuds of an aristocratic society can be seen from two examples. After the
conquest of the fortification Jahna, the Saxons killed all the adults, and took off
the young boys and girls into captivity, meaning slavery. In Merseburg Henry
installed a legion formed of thieves and robbers. He remitted their punishment
and gave them the instruction to undertake as many plundering expeditions
against the Slavs as possible.

We can see a similar difference under Otto tire Great. In October 955, imme-
diately after the victory at the Lechfeld, Otto moved with Saxon contingents
against the Wends under their princeps Stoinef, who by this time was harbouring
the Billung brothers Wichmann and Ekbert, members of the Saxon high
nobility and relatives of the royal house. We know the details of this campaign
from the extensive account given by Widukind of Corvey. 24 After die victory,
in which Stoinef was killed, the Slav leader’s head was hacked off and stuck on a
pole on the battlefield; seven hundred prisoners were beheaded. Stoinef’s
adviser’s eyes were gouged out, his tongue was torn from his mouth, and he
was left helpless among the dead. The cruel treatment may be explained in part
by the massacre carried out by a previous Slav raiding-party, which had pre-
ceded Otto the Great’s reprisal expedition. Duke Hermann had ordered the
garrison of a fortification to surrender to its Slav besiegers under Wichmann’s
leadership on condition that the free and their families should have safe-
conduct, and only the slaves should be left to the Slavs. However, during the
departure there were clashes, and the Slavs promptly killed all the males and

23 Widukind, Res gestae Saxonicae 1,38. 24 Ibid. 111,52—5.

282

GERD ALTHOFF

took off the women and children into captivity. But this explains the cruelties
only in part. One must also bear in mind that the normal restraint shown in
dealings with relatives, fellow-members of a caste or people, and Christians,
did not apply to dealings with the Slavs.

Widukind’s account of the negotiations between Otto and Stoinef is given
much space, and it shows that tire campaign aimed not at incorporation but
rather at satisfaction or revenge. The first step was that a Slav embassy
appeared before Otto, and offered that the Slavs would pay tribute as usual, but
would exercise lordship in their lands themselves; otherwise they would fight.
Otto answered that he was denying them peace only because they were not pre-
pared to render satisfaction for their misdeed. Then he laid waste his oppo-
nents’ territory, but not without in turn sending an embassy to Stoinef, led by
no less a person than Margrave Gero. Otto’s offer was this: if Stoinef would
submit, Otto would be his friend and not treat him as an enemy. It should be
stressed that here a pact of friendship was being offered to a heathen Slav prin-
ceps. The friendship was not to eliminate all traces of lordship: a surrender, a
deditio was to precede the grant of friendship. Yet such behaviour does not
suggest a plan of conquest . 25

That rituals of friendship and treacherous cruelty were found side by side in
the conflicts between Saxons and Elbe Slavs can also be demonstrated from
Gero’s ‘friendship banquet’, in which he had thirty Slav princes killed while
drunk. Widukind excuses this act with the argument that Gero was here merely
anticipating a trick intended by the Slavs themselves, who planned Gero’s
death . 26 Quite apart from the question of whether these accusations are jus-
tified, it is noteworthy that the rituals of peace and friendship also had a place
alongside bitter conflicts. The boundary conditions of these conflicts included
situations in which Gero could invite thirty Slav princes to a convivium or Otto I
become the amicus of the Christian duke of the Poles, Miesco I, and make a
similar offer to the pagan Stoinef.

Nevertheless, contacts between the leading strata of the Slavs and the
Saxons were, taken as a whole, very limited. Only for a few Slav magnates were
Christianity and the Saxon alliance attractive enough for them to surrender
their traditional beliefs and their local ties. One rare example is the Hevellic
prince Tugumir, held in prison by the Saxons since die time of Henry I. He
finally allowed himself to be bribed to betray his own lordship to the Saxons;
the entry of his death in the necrology of Mollenbeck shows that he must have
become a Christian . 27 Pretending to have fled from prison, he returned home,
had his nephew and rival for lordship treacherously killed, and then submitted
himself and his people to Otto’s lordship. According to Widukind, his example

25 Ibid, hi, 5 3—5. 26 Ibid, n, 20.

27 Das Nekrolog von Mollenbeck, ed. Schrader, p. 355 (17 May).

Saxony and the Elbe Slavs in the tenth century 283

was followed by all the Slav tribes up to the Oder, who ‘were ready to pay
tribute’. 28 Here again we must stress the distinction between payments of
tribute and the associations evoked by die word incorporation. Widukind’s
account makes very clear that die Slav tribes were generally prepared to pay
tribute, but die question was whether they would have accepted more wide-
ranging restrictions on dieir freedom.

Apart from Tugumir, we hear of only a small number of cases of collabora-
tion with the Saxons. The normal function of marriages in strengdiening
peaceful ties was here ruled out by the impossibility of marriages between
pagans and Christians. We know nothing about the background to Otto’s
liaison with a noble Slav woman, the modier of die later archbishop of Mainz,
which could take us further here. Only once do we hear of such a marriage plan
in Adam of Bremen’s account, which he explicidy describes as based on oral
tradition. 29 A Slav duke is said to have sought die Saxon duke Bernard I’s niece
for his son in marriage, and this was agreed to. The Wend in return provided a
thousand mounted soldiers for Duke Bernard’s contingent on the Italian expe-
dition, where they almost all fell. Neverdieless, Margrave Dietrich frustrated
the planned marriage with the remark that the duke’s kinswoman was not to be
given to a dog. This, according to Adam (and to Helmold of Bosau, who
follows him), was the reason behind the great Slav uprising of 983. It was cer-
tainly not the only reason, but die anecdote demonstrates the Saxon attitude
very clearly; at all events the peace-bringing function of marriage was litde seen
in relations between Saxons and Slavs. The unnamed Slav duke may well have
been the Abodrite Mistui, who according to Thietmar had a capellanus Avico in
his entourage, 30 and so presumably was not wholly opposed to Christianity,
which would certainly have favoured the marriage proposals. If our fragments
of information do not deceive us, then participation in the uprising of 983 and
a receptive attitude to Christianity were not mutually exclusive; certainly Mistui
appeared only a year later at an assembly held by Henry the Quarrelsome in
Quedlinburg.

In assessing the nature of these conflicts we may further adduce those
Saxon magnates who fought on the side of the Slavs. These included not only
the Billung brothers Wichmann and Ekbert, already mentioned, but also a
Saxon noble named Kizo who went over to the Liutizi out of annoyance with
Margrave Dietrich. It has been argued that the inclitus miles Kizo (Christian),
who came from die Merseburg region, was a member of Margrave Gero’s
kindred. This would show a remarkable similarity of behaviour among
members of the two leading margraval families, Billungs and Geronids: in
each case relatives of die margrave go over to the Slavs as a result of internal

28 ‘tributis regalibus se subiugarunt’: Widukind, Res gestae Saxonicae 11, 21 .

29 Adam of Bremen, Gesta 11, 42. 30 Thietmar, Chronicon iv, 2.

284

GERD ALTHOFF

Saxon conflicts, because they feel that they are not able to get justice in
Saxony.

Still more significant is the fact that these magnates were entrusted by the
Slavs with important command positions. Wichmann led Slav attacks against
Saxony and against Miesco I of Poland; Kizo was in charge of the garrison in
the Brandenburg, whence he also led attacks against Saxony. Even more inter-
esting than the trust and esteem shown by tire Slavs towards these ‘traitors’ is
die fact that diey were not condemned by other Saxon magnates for their beha-
viour. Wichmann and Ekbert were able to flee to Hugh of Francia after the
massacre of Stoinef’s army in 955, until no less a person than Archbishop
Brun of Cologne secured Ekbert’s pardon from Otto the Great. We can
observe this member of the Billung family enjoying unrestricted rights of
political action in Saxony in the period which followed. Wichmann, by con-
trast, who like his brother had previously been declared a ‘public enemy’, 31 was
reconciled with Otto by Margrave Gero. Even after he once again broke the
reconciliation oath he had taken, Gero allowed him to go off once more to the
Slavs, because he saw that Wichmann was in fact guilty. In other words, Gero
had deliberately prevented Wichmann’s condemnation. Kizo offers a similar
story. After he had again changed sides in 993 and submitted himself and the
Brandenburg to Otto Ill’s lordship, he was allowed to retain his position and
given support in his defence against the Liutizi. It was one of his vassals, with
the Slav name Boliliut, who made himself master of the fortification during
Kizo’s absence; Kizo was killed trying to reconquer his position. Thietmar of
Merseburg, who describes the episode at length, makes no criticism of Kizo’s
behaviour; he explicitly allows him the right to change sides ‘to come to his
right’. 32

All details of the border wars as revealed by Widukind and Thietmar point in
the same direction: the wars were an exclusively Saxon affair, concerned with
submission - recognition of overlordship and payment of tribute - but equally
often with revenge, for every attack demanded reprisals. Opponents were
treated savagely and cruelly, often treacherously; but alongside this we also find
negotiations, offers and conclusions of peace agreements, rituals of solidarity
and so forth, which show that all the forms of conflict familiar from the early
middle ages were practised here. The only thing we do not hear of explicitly in
tire narrative sources is any plan to incorporate the Elbe Slavs into the
Ottoman Reich, or of measures taken by die kings to do so.

31 Widukind, Res gestae Saxonicae in, 60. 32 Thietmar, Chronicon iv, 22.

Saxony and the Elbe Slavs in the tenth century 2 8 5

Organisation: legates, margraves, burgwards, bishoprics

As was said at the opening of this section, it has been customary even in recent
scholarship to talk of the ‘introduction of a strict military, administrative and
ecclesiastical organisation’ intended by Otto the Great to produce a ‘definitive
subjugation and incorporation into the Reich’ of the districts east of die
Elbe. 33 The sources offer no specific account of when these organisational
innovations were introduced; the conventional view has been derived from
various indications offered by the sources, which require a critical examination.

The first margraves we meet in the tenth century are Hermann Billung and
Gero, both appointed by Otto the Great in 936 and 937, though here not yet
called margrave. The Saxon reaction to these appointments shows drat die
offices were not new ones; indeed, relatives of the appointees, the Billung
Wichmann the Elder and Otto’s own half-brodier Thankmar, displayed irrita-
tion at having been passed over for positions to which they felt they had greater
claims. The two new office-holders succeeded the Saxon magnates Bernard
and Siegfried, whose position is described in the sources as a legatio. This is
probably to be understood as a military command in a border region, as exer-
cised for example by Bernard at the batde of Lenzen in 929. Besides legatio
(legatus) Widukind of Corvey 34 also used the term princeps militiae to describe
Hermann Billung’s position. Until the death of these two ‘margraves’
Hermann and Gero, in 973 and 965 respectively, both narrative and charter
sources use a whole range of titles for them, among others comes , marchio, dux,
dux et marchio, but there is no reason to suppose that royal reorganisation had
changed in any way the tasks assigned to these magnates. In other words, the
sources do not suggest that Otto introduced a ‘margraval organisation’. We
have absolutely no information about any powers the two may have possessed
over and beyond their military commands - whether, for example, they also
exercised jurisdiction or command over other counts in the border regions.
The two ‘officials’ gained their high profile exclusively through military activ-
ities, which they evidently carried out with a high degree of independence.

There is also considerable uncertainty about what happened to Gero’s
sphere of office after his death. We find no fewer than six counts from the area
under Gero’s command with the title of marchio in the period following his
death, a fact very difficult to interpret. The sources give no indication of why a
single successor to Gero was not appointed; but to deduce that the situation
was now so secure that a single leader was no longer needed is hardly plausible
in view of the uprising of 983. We do not even know whether the various mar-
graves were appointed by tire king or not. Even after tire deaths of the various

33 Ludat (1968/82), p. 46. 34 Widukind, Resgestae Saxonicae u, 4.

286

GERD ALTHOFF

marchiones we cannot observe the appointment of successors. Rather, we find
from the beginning of the eleventh century that three large margravates had
evolved out of the area formerly under Gero’s lordship: the North march, the
East march and the march of Meissen. The transition phase following Gero’s
death does not therefore suggest that precise regions of office had been laid
down in a royal plan which would have allowed a smooth succession following
the death of an office-holder.

The military independence of the margraves also argues against a carefully
worked out royal plan of expansion. Even substantial campaigns were not nec-
essarily cleared with the king in advance, as we can see in 972, when Margrave
Hodo attacked the Polish duke Miesco, even though die latter was amicus impe-
ratoris and a tribute-payer. Otto the Great used messengers to direaten both
participants in the conflict with the withdrawal of his grace, if they should not
keep the peace until his return from Italy. Around the same time the margraves
Hermann and Dietrich made peace with the Redarii and kept it even when
Otto sent written orders from Italy to the contrary. 35 These incidents hardly
speak for royal organisation and planning of activities east of the Elbe; they
tend rather to support the view arrived at in the previous section of local forces
taking ad /wand repressive measures.

Interestingly enough, one of the most ‘well known’ of these independent
actions probably derives from a misunderstanding by Thietmar of Merseburg.
He describes how Margrave Gero subjugated the Polish duke Miesco and his
followers to imperial ditto? 6 This statement, which scholars have discussed at
length and controversially, is in all probability a result of Thietmar’s misreading
of Widukind. 37 Thietmar simply summarises Widukind’s cc. 66—8 in two sen-
tences: Gero subjugated the Lausitz and Miesco; Hermann Billung Selibur and
Mistui. But Widukind writes in c. 66 that Gero returned Wichmann to the Slavs
to save him from being condemned, and Wichmann twice defeated Miesco; his
formulation makes it easy to confuse Gero and Wichmann. There is no reason
to suppose that Thietmar is here drawing on his own knowledge of a campaign
by Gero against Miesco, and this may be struck from the record.

The origins, powers and tasks of the margravates are thus noticeably more
complex than the picture offered by previous scholarship. This is even more
true of die assumption that Otto the Great introduced a burgward organisation
after die marches had been set up. Certainly, the word burgwardium or burgwar-
dum is found in the sources from the middle of the tenth century onwards.
Certainly, fortifications had a central function for the population living in their
vicinity; in times of need the population could take refuge there and they were
also obliged to perform services and make renders. Naturally, such an organ-

35 Ibid in, -jo. 36 Thietmar, Chronicon n. 14. 37 Widukind, Res gestae Saxonicae hi, 66.

Saxony and the Elbe Slavs in the tenth century 287

isation was particularly important in the marcher regions, so that the map
drawn by Walter Schlesinger of the burgivards shows a concentration which is
impressive but hardly surprising. The question is rather how we are to visualise
the organisational measures taken by Otto the Great to install the burgivard
system as a miniature version of die entire Ottoman state structure. 38 From
which groups were the garrisons recruited? Where did the commandants come
from? Who determined the estates and services for the maintenance of the
warriors? Such a bundle of organisational measures would have required con-
siderable activity and the participation of significant numbers of people, most
prominently the members of the Saxon nobility, in their planning and execu-
tion. But such plans have left absolutely no traces in the sources. Moreover, the
fact diat Otto donated several of these burgivards to die newly founded bishop-
rics does not suggest that diere was any kind of strict organisation for the
purpose of expansion; it would have been hollowed out as soon as it was intro-
duced had that been the case.

By contrast we do find organisation and planning in another aspect of
eastern politics, which must be set against the facts sketched so far: ecclesiasti-
cal organisation. Scholarship has tended to see Otto’s missionary and church
policy as a part of his expansion policy. Ottoman missionary activity cannot be
treated as a whole in a chapter devoted to Saxony; but there is no doubt that it is
the history of the episcopal foundations which provides the most detailed
knowledge of property-holding and lordship in the regions east of the Elbe.
The foundation charters for Havelberg and Brandenburg show that Otto the
Great was able to transfer dvitates and tithes in regions of Slav settlement to the
new churches. Later gifts to Magdeburg confirm the impression that the ruler
disposed of a whole range of possessions and rights east of the Elbe; that, in
other words, the idea of state boundaries in our modern sense is quite anach-
ronistic for this period. But equally anachronistic would be any impression
gained from such information that these rights and possessions were an index
of the success of Ottoman expansion. To deduce from the ability to found and
endow bishoprics in die regions of the Elbe Slavs that there must have been an
intention to expand will not work; at precisely die same time (948) three
Danish missionary bishoprics were founded as suffragans of Hamburg-
Bremen, but no one has deduced similar expansionary intentions towards die
Danish kingdom from these foundations. Because Charles the Great had an
interlinked strategy of mission and conquest in his Saxon wars, we do not have
to assume that this must also have been so in the Ottoman period.

There are a number of indications in the sources that ecclesiastical and mili-
tary activities were not coordinated with each other, rather that the forces

38 Cf. Schlesinger (1937/61).

GERD ALTHOFF

involved frequendy blocked and disturbed each other’s activities. Here we may
mention the energetic resistance by die bishop of Halberstadt and tire arch*
bishop of Mainz to die plan to erect an archbishopric in Magdeburg. We may
also mention Otto’s stern warning from Italy to die Margraves Wigger,
Wigbert and Gunther that they should obey the instructions of die new arch*
bishop and furthermore should endow the new bishops of Zeitz, Merseburg
and Meissen adequately, so that these should not be taken to be poor peasants.
The fact that this warning was evidently much needed gives a deep insight into
the contemporary situation, and it fits in well with later ecclesiastical com-
plaints that it was the cruelty and greed of the Saxon margraves which pre-
vented missionary successes and in the last resort was responsible for the great
Slav uprising of 983. The foundation of the archbishopric of Magdeburg was
certainly part of an organisational conception due in essence to Otto I and
realised by him in Italy. What are lacking are convincing demonstrations that
dais conception was only a part of a greater plan of expansion.

SAXON HISTORIOGRAPHY IN THE TENTH CENTURY

The works of so-called ‘Ottoman’ historiography, to which we owe most of
our knowledge of die problems discussed so far, were almost all written in
Saxony, the core region of the Ottoman Reich. To characterise them as
‘Ottoman’ implies that the works were written from the perspective of the
king. But this assessment conceals essential characteristics of these works.
More recently there has been a strong tendency to ask how far they witness to
others’ opinions and positions, not identical with positions of the ruling house.
The idea of an Ottoman ruling house in any case plays down the divergent
forces operating within this ‘house’. The new view of ‘Ottoman’ historio-
graphy proceeds from the observation that almost all works of Ottoman
historiography were written within ecclesiastical communities - Corvey,
Quedlinburg, Gandersheim, Nordhausen, Merseburg - and often at crucial
phases in the development of these communities. We are thus confronted with
a basic question about the function of such historical writing, which is prob-
ably much more linked with and aimed at influencing its own present than one
normally assumes of historical writing. In Ottoman Saxony, historiography
allowed forces to articulate themselves whose opinions and interests were very
different from those held by the Ottoman rulers.

The anonymous Lives of the Ottoman Queen Matilda, written in her foun-
dation of Nordhausen, provide a very specific view of the history of the
Ottoman dynasty. The older work, written around 974, is aimed at Otto II; the
younger was intended for Henry II, which led to the rewriting of important
passages. Henry II’s direct ancestors were transformed in the younger Life into

Saxony and the Elbe Slavs in the tenth century

289

central figures in Matilda’s life. Her whole love and care was devoted to diem,
as were her mourning and her memory. There is thus a remarkable adjusting of
historical writing to accommodate its addressee. But both Lives have a central
theme in common: Matilda devoted all her energy to ensuring that the future of
her foundation of Nordhausen should be secured by all kinds of legal protec-
tion. She did this, allegedly, with the understanding and close cooperation of
her son Otto the Great. For - and this is also a central message of the works -
he had learned that the preservation and well-being of his rule depended in
decisive measure on whether he supported his mother’s attempts to found and
endow ecclesiastical communities using her dower lands. Not until his rule met
with failures and crises and he had been warned by his wife Edith did the king
change his behaviour towards his mo ther.

This remarkable pointing up of Ottoman family history, accentuating the
discord between mother and son with all its consequences, becomes compre-
hensible only when one realises that the two Lives were written at a point when
new queens were about to receive their dower. In the case of Otto II and his
wife Theophanu, the dower charter has survived in the form of the famous
purple dowry charter . 39 In this we read that Theophanu’s dos includes
Nordhausen, and it is stated explicitly that this means everything which Queen
Matilda possessed there. This threatened tire monastic community in
Nordhausen, should the new queen decide to disturb Matilda’s work. To
prevent this the Nordhausen community composed the Life of Matilda and
dedicated it to Otto II. After the warning implicit in the Life had been success-
ful, the procedure was repeated when Queen Kunigunde, wife of Henry II,
came to die throne. In other words, a spiritual community used spiritual means
in a position of existential threat by depicting its founder’s life as an exemplum
for the new queen, and it reinforced this admonition with the clear warning
that action contrary to Matilda’s intentions would bring down God’s anger and
punishment.

The depiction of the early history of Gandersheim by tire nun-poet
Hrotsvitha starts with a redirected Annunciation. It was prophesied by no less
a person than John the Baptist to Aeda, die mother of the foundress Oda, that
her seed should found a monastery, Gandersheim. This would ensure the
peace of the Reich, ‘as long as its vows are protected by the kings’ care’. As a
reward for its foundation the family would receive so high a dignity ‘that no
other of the kings on earth would dare to place himself alongside it in rank and
powerful majesty ’. 40 This is of course a retrospective prophecy; but we
must ask why Hrotsvitha linked the well-being of the Ottoman house so expli-
citly with the furthering of Gandersheim. Moreover, this theme shapes tire

39 D O II 21 . 40 Hrotsvitha, Primordia coenobii Ganders heimensis, ed. Homeyer, p. 452.

290

GERD ALTHOFF

subsequent account in the Primordia. The members of the Liudolfing house are
implored to do everything necessary to secure the safety, the protection and the
material endowment of the monastery. The account lays particular stress on
the appropriate actions of Duke Liudolf and his wife Oda and those of Duke
Otto, all of whom, according to Hrotsvitha, knew very well just how much the
success of their house depended on the merits and prayers of the
Gandersheim nuns. Just as in Nordhausen, so in Gandersheim a clear warning
was issued to die ruler in die form of a historiographical work which tried to
oblige him to follow the example of his ancestors. Hrotsvitha’s second work,
the Gesta Ottonis, should also be seen in the context of this warning. It was
written at the request of Abbess Gerberga and interestingly enough was to be
laid before Archbishop William of Mainz, Otto’s chief opponent in the
Magdeburg question, for his approval. Unfortunately, the fragmentary nature
of the work makes the author’s intentions ultimately unclear, but one thing is
certain: die Gesta stress die internal crises of Otto’s rule and his relatives’ upris-
ings against him. They also depict his opponents very positively, and they stress
repeatedly that it was God’s grace alone which rescued Otto from great peril
and preserved his rule. If Otto is again and again compared with David in this
context it must be asked whether the comparisons are intended to be praising
or warning.

The most famous work of ‘Ottoman’ historiography, Widukind’s Saxon
History , was dedicated in 967 or 968 to Matilda, abbess of Quedlinburg and
daughter of Otto I. Although each of the three books of the Saxon History
begins with a prologue dedicated to Matilda, scholarship has largely ignored
die question of why such a work should have been dedicated to an imperial
daughter, and why this should have happened in the years 967-8. The historical
context of the dedication offers a number of clues. The young Matilda (she was
eleven years old) had been made abbess of Quedlinburg at Otto’s request by all
the archbishops and bishops of the kingdom at a great festival in April 966 at
Quedlinburg, before the emperor set out for his third Italian expedition. In the
autumn of 967 his son and co-ruler Otto followed him south. The only
members of the royal house remaining north of the Alps were William of
Mainz and Queen Matilda, widi William acting as regent. The old queen fell
seriously ill at the beginning of 968. William hurried to her sick bed, and must
have realised that her death was imminent, but surprisingly he himself died
before the queen at the beginning of March; she followed him on March 14.
The young Abbess Matilda was thus tire only member of die royal house left
nordi of the Alps, and this remained die case for four years. She now had the
task of representing Ottoman rule in Saxony in a very difficult situation. Otto
die Great ordained the foundation of the archbishopric of Magdeburg and its
suffragans of Merseburg, Zeitz and Meissen from Italy; up to the last minute

Saxony and the Elbe Slavs in the tenth century 291

there had been serious opposition to these plans in Saxony. It was precisely in
this situation that Widukind dedicated his Saxon History to the abbess; it con-
tained just what she needed in her current position, the knowledge of the past
required to be able to rule in Saxony. This was the precise function of die Saxon
History. It allowed Matilda to act as die highest-ranking person in Saxony, as the
domina imperialism because she had been informed about the history of the
Saxon people, about her father’s and grandfather’s achievements, and not least
because she knew about the difficulties which her father’s decisions, in Saxony
in particular, had brought about. For diis purpose Widukind included a
number of pieces of information really belonging to the secreta regis, to that area
which a historiographer would normally have discreetly passed over. One must
read Widukind’s work in the light of die situation in which it was composed in
order to understand why it is precisely diose things referred to over and over
and not others which are dealt with; it was these which Matilda had to know
about in order to act independently and as a member of the imperial house in
Saxony from 968 onwards.

The Quedlinburg annals are equally shaped by their situation. They were
written in Henry II ’s reign, at a time when Quedlinburg lost its former domi-
nant role as a royal centre. Among other things, Henry broke with the Ottoman
tradition of celebrating Easter there and hardly visited the place at all, which
evidently wounded Quedlinburg sensibilities deeply. These were expressed by
the annals in their account of the year 936: Queen Matilda intended to make
Quedlinburg a ‘kingdom for the gentiles’ and had therefore collected only well-
born persons, for these seldom go astray. 41 The term regnum gentibus is evidently
an echo of the tide rex gentium, ‘king of the peoples’, used by Widukind for
Otto I at the end of his Saxon history. 42 How did this community react to its
downgrading by Henry II? None of his actions finds a good word. The annals
from 1003 are a drastic demonstration of how openly criticism of a ruler might
be practised in a royal monastery. A few examples: ‘The king, very down-
hearted because he had won no good peace, returned with a miserable army
and brought the bodies of the dead with him’ (1005); 43 ‘As the king learned of
this he was troubled in his heart and enjoined his men not to leave the matter
unavenged. But - I know not for what reason - up to the present so great an
anger has not been turned into deeds’ (1007). 44 When Henry II chose a differ-
ent candidate for the archbishopric of Hamburg-Bremen not the one elected
by the chapter, the annalist commented on the events as follows: ‘But the king’s
crude ways/ and his thirst for gain, thrust die petitioner back,/ turn from the
weeping his gaze.’ 45 Henry’s Roman expedition of 1014 is summed up thus:

41 Annales Quedlinburgenses, s.a. 936, p. 54. 42 Widukind, Res gestae Saxonicae hi, 76.

43 A nnales Quedlinburgenses, s.a. 1005, p. 78. 44 A nnales Quedlinburgenses, s.a. 1007, p. 79.

45 Annales Quedlinburgenses, s.a. 1013, p. 81.

292 GERD ALTHOFF

‘After he had . . . ordained official matters well, as he thought, and collected
huge sums of gold from everywhere, he accelerated his return homewards,
though not without inflicting damage on many people’. 46 Such unrestrained
criticism of die ruler corresponded completely to the contempt which Henry
had showed towards Quedlinburg by staying in Saxony frequently enough but
never visiting the place, and by making it no gifts until 1 02 1 . It is hardly possible
for a community to lose the king’s grace and presence more quickly than
Quedlinburg did. Seen in the context of other changes which Henry’s reign
brought with it, such an observation also shows how far he broke with die tra-
dition of his Ottoman ancestors, indeed distanced himself from them. The
loss of royal favour did not last for the whole of Henry’s reign, however. In
1014 he granted Abbess Adelaide of Quedlinburg the headship of die nunner-
ies of Gernrode and Vreden. In 1021, at last, he attended the consecration of
the newly built monastic church at Quedlinburg and made a rich gift to the
convent. It can thus hardly be coincidence that the negative comments on
Henry II cease from 1014 onwards and that from 1021 the author is once again
capable of panegyric descriptions of Henry II’s deeds of a kind familiar to us
from the early years of Ottoman rule. Historical writing thus reacted directly to
changes in die political climate.

To ask about the cause and historical context for die origins of ‘Ottoman’
historiography is to sharpen one’s perception for the specific functions of each
of these works; none was written exclusively from the king’s perspective, and
some were written directly contrary to it. Forces widiin Saxony with interests
not unconditionally identical with those of the rulers could articulate diem-
selves in this way: ecclesiastical communities like Nordhausen, Gandersheim
and Quedlinburg, and then a littie later Thietmar of Merseburg, who wrote the
history of his precarious and threatened bishopric in such a way that it can also
be written as a history of die Reich and the Saxon people. Widukind of Corvey
wrote not so much as a representative of the oldest monastery in Saxony as of
those forces within Saxony who wished to avoid conflicts, and so he took care
to instruct the young imperial daughter about positions and patterns of beha-
viour of which her father would certainly not have approved. Saxon historiog-
raphy of the tendi century can thus hardly be characterised adequately as
‘Ottoman’ historiography, still less as ‘Ottoman house tradition’. It should
radier be seen as formulating the perspectives of forces within Saxony which
were certainly not fundamentally opposed to Ottoman kingship but were
equally not prepared to identify themselves unreservedly with all the positions
and decisions taken by these kings.

46 Annales Quedlinburgenses, s.a. 1014, p. 82.

CHAPTER I I

BAVARIA IN THE TENTH AND EARLY
ELEVENTH CENTURIES

Herwig Wolfram

THE CAROLINGIAN INHERITANCE

The restoration and expansion of the Frankish empire in the eighth century
were possible not least because the ruling Carolingians accepted the existence
of the regnal structure of the Franco-Lombard core region of Europe and
indeed took this on board as a permanent aspect of Carolingian tradition. The
sources distinguish between three kinds of regna. In its first sense, regnum means
the whole Carolingian empire; in its second sense it refers to a Frankish (or the
Lombard-Italian) subkingdom; and in die third it denotes a political and
regional entity with a name drawn from that of a people living there under a
common law The first two kinds of regna were invariably ruled by kings,
whereas a regnum of the third type might be ruled by kings’ sons (widi or
without a royal title) or by princes without kingly rank. The Carolingian empire
was thus a fiexible polity built up of prefabricated parts, an organisational form
which allowed an imperial extensiveness together with a governmental inten-
siveness in smaller regions.

The Carolingians had in general to associate their leading men with the
government of the empire; die ‘imperial aristocracy’, to use Gerd Tellenbach’s
term, were still more entided to political participation in die regna of the diird
type. Representatives of the most successful aristocratic groupings emerged
from the competition for closeness to the king, power and influence as the
‘second after the king’. Diplomata and other sources written in the royal
entourage never call such magnates anything except comites or, in die late
Carolingian era, marchiones. A secundus a rege of this kind might neverdieless
acquire a princely position, even in die Carolingian period. Occasionally such a
comital or margraval office-holder might even have to take on royal duties, if
royal audiority had failed or been withdrawn from an area for one reason or
another. Thus Odo of Paris and other non-Carolingian ‘princes of the Franks’
were elected as kings in 888 because on Charles Ill’s death his regna ‘were
deprived of their natural lord’. 1

1 Regino, Chronicon, s.a. 888.

293

HERWIG WOLFRAM

294

One of the most prominent regna of the third type was Bavaria, which had
had the status of a regnum even in pre-Carolingian times and had consolidated
this position in tire course of the ninth century. In the summer of 8 14 Louis the
Pious had organised tire Frankish subkingdoms, among which Bavaria was
named for the first time. Three years later he issued tire ordinatio imperii. Louis’
namesake, known to historians under a tide whose inaccuracy is sanctioned by
tradition as ‘Louis the German’, received as his kingdom ‘Bavaria, the
Carinthians, the Bohemians and the Avars, as well as the Slavs who live in the
east of Bavaria’. 2 In odier words, Louis became king of two regionally defined
regna. die Bavaria which had been ruled over by the Agilolfing duke Tassilo III
and the peoples of the Bavarian Easdand, which included semi-autonomous
Slav peoples and the dependent Avar khaganate on the middle Danube.
Between 830 and 833 Louis the German even used the tide rex Baioariorum in
the charters he issued in his own name, so that Bavaria received the highest
form of political recognition for the first time.

In 833 the three adult sons of Louis die Pious rebelled against their father,
and Louis the German took on die regnum in orientali Francia? This ‘kingdom in
eastern Francia’ lay largely east of the Rhine, and corresponded roughly to the
term Germania as used in classical literature. The struggles over Louis die Pious’
inheritance led Louis the German to set up a Bavarian viceroyship, a secundus a
rege (‘second after the king’); he and his immediate successors, however,
remained strong enough to keep the holders of this office in their places.
Kings’ sons and princes established themselves in one of the two parts of the
Bavarian double regnum in accordance with, not against die will of, die east
Frankish king. After Louis the German’s deadi in 876 Bavaria was briefly
reunited under his oldest son Carloman. However, the new king conferred the
Bavarian Easdand on his illegitimate son Arnulf ‘of Carinthia’, in much the
same way as he himself had held it during his father’s lifetime. It was from here
that Arnulf set out with a powerful army of Bavarians and Slavs in 887 via
Regensburg to Frankfurt, where in November of that year he was set up as
king. The regnum in die Bavarian Easdand had grown noticeably in strength.
The person who held it was not just able to take over first parts and then the
whole of old Bavaria, as in Carloman’s time; its possessor could now also
become the heir to the whole of Louis the German’s kingdom, provided that
he moved quickly to take control of the region around the confluence of Rhine
and Main, and so dispose of die most important transport routes of the east
Frankish kingdom and of the royal fisc in Rhenish Franconia.

On 8 December 899 Arnulf, who had been crowned emperor in Rome
nearly four years earlier, died at Regensburg; he was succeeded by Louis ‘the

2 For the terminology see Eggert (i 973).

3 The title used in DD L G 13 and following diplomata.

Bavaria in the tenth and early eleventh centuries 2 9 5

Child’, who was at die time not quite seven years old. In Bavaria a certain
Luitpold seized the initiative and prevented a power vacuum from arising, dius
showing both fidelity to the Carolingian house to which he was related and an
eye to the main chance. Luitpold, whose family is known to historians as die
‘Luitpoldings’, had to share power with other magnates, but stood at their
head. Already in the year which followed Arnulf’s death Luitpold began cam-
paigning against die Hungarians on his own initiative and erecting fortifica-
tions against them. In the summer of 907 he took the offensive against diese
enemies, who had already conquered a great part of the Bavarian east; on 4 July
907 he lost die batde, a Bavarian army and his life near Bratislava on the
Danube. Bavarian infiuence on the king and the rest of the east Frankish
kingdom declined sharply after the batde; Louis the Child shifted his principal
residence from Regensburg to Frankfurt. The Carolingian period had already
come to an end in Bavaria before the east Frankish royal house died out on the
death of Louis the Child on 24 September 9 1 1 .

Like other similar magnates in the odier regna of die east Frankish kingdom,
Luitpold had achieved the status of a prince in Bavaria. In spite of his catas-
trophic failure as leader of the Bavarian army, Luitpold’s position was so strong
that his son Arnulf was able not only to succeed him but to rise to the rank of a
quasi-royal prince of die Bavarians, a dux Bavvariorum, a basis from which he
too could undertake a ‘new start on the basis of Carolingian tradition’. 4 This
tradition included a rich variety of possibilities, contradictions and challenges.
It corresponded to Carolingian tradition diat the rulers of Bavaria were to
intervene in Italian affairs even before those of Alemannia did so, and even to
seek the imperial crown, just as Carloman, the oldest son of Louis the German,
had tried to do and as his son Arnulf was to succeed in doing. But Louis the
German had also given his second son Louis a kingdom which in 876 had
already united Franconia and Saxony and provided a basis for possession of
Lotliaringia, a decision which anticipated the political situation under Conrad I
and Henry I. A further aspect of Carolingian tradition was that of the
unanointed king in east Francia, whereas the other kings in the regna of the
second type had generally been anointed. So also were die setdements between
these regna on the basis of treaties of friendship, which the kings made with
each other for the preservation of peace; in addition to these, die Carolingian
feudal system came to establish ‘interregnal’ ties. Avar, Slav, Lombard and
Breton princes became the men of Carolingian kings. The end of the ninth
century also saw vassalitic treaties within the Frankish core lands which
enabled a Carolingian ruler to establish himself as high king, as primus inter pares
among his own kin or non-Carolingian competitors alike. Rule over the church

4 Cf. the title of Althoff and Keller ( 1985 ).

296

HERWIG WOLFRAM

was a self-evident part of Carolingian kingship. The episcopate preferred on
the whole to be a part of a supra-regional regnal church rather than of a church
confined to a regional regnum of the diird type. The four older archbishoprics in
the east Frankish kingdom were here concerned not least with their own rank
in die court chapel and the royal chancery. Mainz stood for Franco-Saxon
unity, while from the time of Carloman’s rule onwards Salzburg could
be certain of pre-eminence within Bavaria or any east Frankish kingdom
dominated by Bavaria. Trier and Cologne disputed pre-eminence within
Lotharingia.

Tensions between norm and reality continued in existence in Bavaria. To
take only two key examples: it was the king, according to Bavarian law, who
nominated the duke, and yet as late as 1002 King Flenry II could refer to the
Bavarians’ right of election, which forbade him an independent decision . 5
Already in the Agilolfing period the dukes had governed the church, and yet
the Bavarian law contained a provision that the king installed die bishop and
that the people elected him . 6 Counts and counties, feudalism and royal vassals,
military and court service were equally part of tire Carolingian tradition in the
various regna, and were to be particularly significant in regions like Bavaria,
where such institutions had merged with or overlaid native traditions. ‘Comital
organisation proved to be one of the most essential instruments of royal
government of the kingdom, the fundamental organisational unit of the
kingdom in matters of administration, justice, and the raising of armies ’. 7 The
beginnings of Carolingian feudalism are to be found in Bavaria in die late
Agilolfing era. There were Bavarian vassals of the Frankish rulers both within
and beyond the regnum. By the second half of the ninth century at the latest the
relations between the east Frankish king and the Slav princes were also organ-
ised on feudo-vassalitic lines.

A further tradition of Carolingian Bavaria was its polyethnic structures
based on Roman, German and Slav traditions, and its openness to the south
and west. Bavaria was the only east Frankish regnum which had frontiers on to
both the Slavic world and the areas of Romance speech. It was Bavarians who
were the first to be termed Nemci by their Slav neighbours, and they were to be
termed ‘Germans’ ( Theotisci ) by the Lombards of northern Italy long before
this name came to denote all ‘Germans ’. 8

Among the Slav neighbours of Carolingian Bavaria the Moravians had
formed die most powerful polity both in political and in ecclesiastical terms. It
had not been possible to conquer the Moravian kingdom; all attempts to treat the
successors to the Avars north of the Danube in the same way as the Bohemian,
Pannonian or Dalmatian Slavs had failed. Louis the German’s counts were left to

5 Thietmar, Chroniconv, 14. 6 Lex Baiwariorum 1,10. 7 SchuLze (1973), p. 347.

8 Wolfram (1991).

Bavaria in the tenth and early eleventh centuries 297

master the situation using their own resources; yet all those who tried to establish
peace on tire middle Danube with the help of negotiations and treaties sooner
or later came to seem like rebels in the eyes of the east Frankish rulers.
Numerous counts lost their positions, but the king’s son Carloman and the
comital generation which succeeded the ‘unfaithful’ ones did not do any better.
Bloody and bitter campaigns alternated with peace agreements until the
Hungarians destroyed the Moravian kingdom, by 906 at the latest.

The ‘Avars, who are now called Hungarians’, 9 had thus come to replace their
predecessors, which meant for the Bavarians that they were to be combatted
with all the means available, but also that one sat down at the same table with
them, and that not just in order to murder them treacherously but also to nego-
tiate honestly and conclude treaties with them. In this way it was possible for
Duke Arnulf, who had defeated the Hungarians three times in 909, 910 and
913, to go into exile among them in 914 and then again in 916, which presup-
poses the existence of a peace agreement at the very least. The same holds true
for 937, when the Hungarians ‘peacefully’ marched through Bavaria on their
way west. The phrase ‘the raging sword of the heathen’ was used on the
Danube as well as elsewhere, 10 but it was only in die distant hinterland that the
Hungarian was seen as a ‘Scythian’, as a member of a people whose name
called forth apocalyptic associations which, significantly, were being noted for
the first time west of the Rhine in this period.

The Bavarian view of the treaty which Charles III had made with the
Nordimen at Asselt in 882 reveals a similarly nuanced view of ‘heathens and
barbarians’. The Mainz version of die Annals of Fulda saw in it the shameful
capitulation of a weak non-ruler advised by a traitor, Liutward of Vercelli; the
Regensburg continuation stressed die friendship between the two sides, who
spent two days in joyous conviviality and underlined the peace treaty with
mutual gifts. 1 1 Not to defame the enemy from the east as the product of hell, or
to do so only half-heartedly, an attitude already adopted by Tassilo III and his
wife Luitpirc, and to treat him as an object of international law, to use modern
terminology, made him into an enemy like any odier; ‘any other’ in this context
might mean a Frankish king like Charles the Great, Conrad I or Henry I. In
odier words, not much distinction was drawn between Avars and Hungarians
on the one hand or an opponent who came from die equally foreign parts of
Franconia and Saxony. As late as 937 the view was taken in Regensburg drat
‘the Saxon Henry had invaded the land of die Bavarians as an enemy’, and that
before him King Conrad I had entered die country ‘not as a king but as an
enemy’. 12 This means that it was not a question of treachery or high treason

9 Annales Fuldenses, s.a. 894, 896, 900, pp. 125, 129, 134.

10 Fragmentum de Arnolfo duce Bavvariae, ed. Reindel. 11 Finn ales Fuldenses, s.a. 882, pp. 98—9, 108—9.

12 ‘Non regaliter sed hostiliter’: Fragmentum deA.rnulfo duce Bavvariae , ed. Reindel.

298

HERWIG WOLFRAM

when Berthold of Reisenburg, the grandson of Duke Arnulf, is said to have
warned the Hungarians of the approach of Otto I and his army before the
batde of tire Lechfeld on 10 August 95 5; it was merely part of an unsuccessful
alternative to die policies successfully pursued by the Saxon Otto which so
impressed both his contemporaries and later observers. It is not surprising diat
these later observers should have attributed a rapid and terrible end to
Berthold, but in reality he was still alive in 976 and was treated respectfully by
Otto II, the son of the ‘betrayed’ king of 955, even though he had evidendy
supported Duke Henry II (the Quarrelsome) and thus once again resisted royal
audiority. The Bavarians of the tenth century neither belonged to a kingdom of
Germany, which did not yet exist, nor were capable of betraying an equally
non-existent German national consciousness.

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "The New Cambridge Medieval History"

See other formats